Jump to content


Match-making totally ruined the game

matchmaking

  • Please log in to reply
57 replies to this topic

pen_pen_1 #41 Posted 15 January 2019 - 03:08 PM

    Private

  • Players
  • 6 battles
  • 8
  • Member since:
    01-12-2019

View PostCA_vampire, on 15 January 2019 - 12:59 PM, said:

 

No, they never said that MM is random. If you want to claim this, then please find this announcement and give us the link. If MM was random, why don't they have that in their web site. 

 

Furthermore, please think a little bit about what "random" means. In the MM there are obviously some rules. Actually, if you think about it, there are A LOT of rules! First of all, the MM has to pick 2 tiers. Which ones? Why 6-7 and not 7-8? How does it pick the two tiers? It is rules! How does it pick how many tanks from 7 and how many tanks from 8? More rules! What type of tanks? Which tanks? Which players? All those are computational rules. It is not just "random", or you could see tier 1 vs tier 10. It is rules and more rules and yet more rules. And we have no idea what the MM rules are. MM is an computational algorithm, a complicated "recipe". It is very naive that people believe it is "just random", biological Evolution works with "random" but that's not how software works.  

 

 

 

 

 

 

 

Moon A orbits the earth once every X days and Moon B orbits the earth once every Y days.

Moon A orbits Mars once every Z days and Moon B orbits Mars once every W days.

 

Newton's analogy between falling objects and moon's orbit. relative mass and the resultant force.

 

the existence of XX%ers, YY%ers and ZZ%ers can neither prove nor disprove the existence of randomness.

 

i don't think you should waste your time to talk to a troll. he is one of the offenders on this forum.


Edited by pen_pen_1, 15 January 2019 - 04:02 PM.


Dan_Deerso #42 Posted 15 January 2019 - 06:41 PM

    Tankologist

  • Players
  • 24224 battles
  • 3,905
  • [III]
  • Member since:
    11-13-2013

View Postpen_pen_1, on 15 January 2019 - 10:08 AM, said:

 

Moon A orbits the earth once every X days and Moon B orbits the earth once every Y days.

Moon A orbits Mars once every Z days and Moon B orbits Mars once every W days.

 

Newton's analogy between falling objects and moon's orbit. relative mass and the resultant force.

 

the existence of XX%ers, YY%ers and ZZ%ers can neither prove nor disprove the existence of randomness.

 

i don't think you should waste your time to talk to a troll. he is one of the offenders on this forum.

 

Did your last alt get banned, too?

__Crusader6__ #43 Posted 16 January 2019 - 01:02 AM

    ACE Chaser

  • Players
  • 58313 battles
  • 11,240
  • [-C6-]
  • Member since:
    12-08-2014

View PostTO_Dominator, on 15 January 2019 - 08:23 AM, said:

 

Apart from matching tank types, platoons, and the tier of vehicles, how is MM not random? 

 

If it's not random, why do 40%, 30% and 60% players exist?

 

And is there ANY reason for MM not to be random? Ask yourself this.

 

WG hates CA_Vampire clearly.  

 

I guess the major issues is it all depends on your definition of random. 

 

 As much as I’m not CA_vampire biggest fan he brings up a lot of good points no Wargaming is not going to divulge their MM algorithms they view that as a trade secret and electro property to be protected at all costs they have told us about the random number generator that it does computations every time Michelle is fired as to dispersion as to penetration in damage leave divulge that I I think we can take them at their word for that    

 

The key notes of what wargaming hasn’t talked about in M&M is open but the waiting criteria for the MM selection it’s Dave said a variety of criteria that it isn’t but haven’t denied that it could be certain other aspects does not truly random is it random enough for the purposes of what we consider potential he I’m obviously not a computer programmer nor mathematician.  


 
Tank Hoarder: 364 tanks in Garage:  365/367 aced (AMX 30B, Mk1 and T49A repo),    wallet warrior.  Loyal M60 owner
I think 5.5 was good for the game - But I also want my Kenny OP nerfed. 
Sorry I haven’t aced the Smasher yet - So I’m still ruining tier 6-8 

 


Absolute_Sniper #44 Posted 16 January 2019 - 01:17 AM

    First Sergeant

  • Players
  • 33362 battles
  • 2,274
  • [CRU2L]
  • Member since:
    11-05-2015

View Post__Crusader6__, on 16 January 2019 - 01:02 AM, said:

 

WG hates CA_Vampire clearly.  

 

I guess the major issues is it all depends on your definition of random. 

 

 As much as I’m not CA_vampire biggest fan he brings up a lot of good points no Wargaming is not going to divulge their MM algorithms they view that as a trade secret and electro property to be protected at all costs they have told us about the random number generator that it does computations every time Michelle is fired as to dispersion as to penetration in damage leave divulge that I I think we can take them at their word for that    

 

The key notes of what wargaming hasn’t talked about in M&M is open but the waiting criteria for the MM selection it’s Dave said a variety of criteria that it isn’t but haven’t denied that it could be certain other aspects does not truly random is it random enough for the purposes of what we consider potential he I’m obviously not a computer programmer nor mathematician.  

Dave and Michelle like M&M’s?



__Crusader6__ #45 Posted 16 January 2019 - 02:36 AM

    ACE Chaser

  • Players
  • 58313 battles
  • 11,240
  • [-C6-]
  • Member since:
    12-08-2014

View PostAbsolute_Sniper, on 15 January 2019 - 08:17 PM, said:

Dave and Michelle like M&M’s?

 

Voice text bit me.

 

 Will fix 


 
Tank Hoarder: 364 tanks in Garage:  365/367 aced (AMX 30B, Mk1 and T49A repo),    wallet warrior.  Loyal M60 owner
I think 5.5 was good for the game - But I also want my Kenny OP nerfed. 
Sorry I haven’t aced the Smasher yet - So I’m still ruining tier 6-8 

 


CA_vampire #46 Posted 16 January 2019 - 07:52 AM

    First Sergeant

  • Players
  • 24268 battles
  • 1,056
  • [MOMB]
  • Member since:
    11-21-2016

View PostDan_Deerso, on 15 January 2019 - 06:40 AM, said:

 

Yes, because there's always enough players available to fill two teams at any given time with zero wait time. I don't have to read your resume, either. From what you say it is obvious to me that you've never worked on anything practical or time-restricted. 

 

 

Since you're lazy, here ya go.

 

 

 

Man, you have reading comprehension problems, too! Perhaps your middle school wasn't very good, but since then it would be good for you if you open a book once a while.

CA_vampire #47 Posted 16 January 2019 - 08:16 AM

    First Sergeant

  • Players
  • 24268 battles
  • 1,056
  • [MOMB]
  • Member since:
    11-21-2016

View Post__Crusader6__, on 15 January 2019 - 06:02 PM, said:

 

WG hates CA_Vampire clearly.  

 

I guess the major issues is it all depends on your definition of random. 

 

 As much as I’m not CA_vampire biggest fan he brings up a lot of good points no Wargaming is not going to divulge their MM algorithms they view that as a trade secret and electro property to be protected at all costs they have told us about the random number generator that it does computations every time Michelle is fired as to dispersion as to penetration in damage leave divulge that I I think we can take them at their word for that    

 

The key notes of what wargaming hasn’t talked about in M&M is open but the waiting criteria for the MM selection it’s Dave said a variety of criteria that it isn’t but haven’t denied that it could be certain other aspects does not truly random is it random enough for the purposes of what we consider potential he I’m obviously not a computer programmer nor mathematician.  

 

I don't think that their main reason is to "guard their secrets".  I don't think that any other company would want to copy the MM we see in Blitz. If nothing else, a lot of people complain about MM. MM is not the most attractive element of this game. 

 

In my opinion, the reason they don't explain how MM exactly works is the same reason they usually don't explain how exactly the events work. This is a (bello)russian company, they give very little (near zero) emphasis to customer service, and to transparency, and to customer relationships. They simply don't care, so they don't hire top talent in customer service and communications. 

 

 



dangerousdan26 #48 Posted 17 January 2019 - 06:44 AM

    Lance-corporal

  • Players
  • 11696 battles
  • 19
  • [ZBLUE]
  • Member since:
    06-14-2018

Matchmaking sucks!

 



_irongiant_ #49 Posted 17 January 2019 - 02:28 PM

    Senior Sergeant

  • Players
  • 32357 battles
  • 582
  • [-DERP]
  • Member since:
    03-02-2016

View PostCA_vampire, on 16 January 2019 - 08:16 AM, said:

 

I don't think that their main reason is to "guard their secrets".  I don't think that any other company would want to copy the MM we see in Blitz. If nothing else, a lot of people complain about MM. MM is not the most attractive element of this game. 

 

In my opinion, the reason they don't explain how MM exactly works is the same reason they usually don't explain how exactly the events work. This is a (bello)russian company, they give very little (near zero) emphasis to customer service, and to transparency, and to customer relationships. They simply don't care, so they don't hire top talent in customer service and communications. 

 

 

 

I agree whole heartedly. There absolutely is no other explanation for their actions.

Bachus131 #50 Posted 17 January 2019 - 02:44 PM

    Sergeant

  • Players
  • 21763 battles
  • 265
  • [JAWZ]
  • Member since:
    12-28-2017

View PostSpartacusDiablo, on 11 January 2019 - 07:33 AM, said:

Bad players are what is ruining the game.

 

Yes this it seems since Mad Games the new meta is to YOLO repeatedly and hope that you get carried.

I was looking at my Churchill III which used to be 65% win rate and despite it's kill  and average damage ratio going up it's win rate is going down.

I have seen even 60%+ players YOLO'ing into 4 enemy tanks so win rate means very little to me atm.


Edited by Bachus131, 17 January 2019 - 02:49 PM.

'Don't follow the lights, they will lead you to your DOOM'  ( Gollum )

__Crusader6__ #51 Posted 17 January 2019 - 05:36 PM

    ACE Chaser

  • Players
  • 58313 battles
  • 11,240
  • [-C6-]
  • Member since:
    12-08-2014

View PostBachus131, on 17 January 2019 - 09:44 AM, said:

 

Yes this it seems since Mad Games the new meta is to YOLO repeatedly and hope that you get carried.

I was looking at my Churchill III which used to be 65% win rate and despite it's kill  and average damage ratio going up it's win rate is going down.

I have seen even 60%+ players YOLO'ing into 4 enemy tanks so win rate means very little to me atm.

 

WR itsel (and especially since 5.5), is misleading.  A lot of players mercilessly clubbed their WR. 

   

   


 
Tank Hoarder: 364 tanks in Garage:  365/367 aced (AMX 30B, Mk1 and T49A repo),    wallet warrior.  Loyal M60 owner
I think 5.5 was good for the game - But I also want my Kenny OP nerfed. 
Sorry I haven’t aced the Smasher yet - So I’m still ruining tier 6-8 

 


__V_O_P__ #52 Posted 17 January 2019 - 09:42 PM

    Sergeant

  • Players
  • 28169 battles
  • 282
  • [CURSD]
  • Member since:
    09-06-2016

View PostCA_vampire, on 16 January 2019 - 08:16 AM, said:

 

I don't think that their main reason is to "guard their secrets".  I don't think that any other company would want to copy the MM we see in Blitz. If nothing else, a lot of people complain about MM. MM is not the most attractive element of this game. 

 

In my opinion, the reason they don't explain how MM exactly works is the same reason they usually don't explain how exactly the events work. This is a (bello)russian company, they give very little (near zero) emphasis to customer service, and to transparency, and to customer relationships. They simply don't care, so they don't hire top talent in customer service and communications. 

 

 

 

Vamp - what would you want MM to be? How would a good MM work?



CA_vampire #53 Posted 17 January 2019 - 10:22 PM

    First Sergeant

  • Players
  • 24268 battles
  • 1,056
  • [MOMB]
  • Member since:
    11-21-2016

__V_O_P__, on 17 January 2019 - 02:42 PM, said:

 

Vamp - what would you want MM to be? How would a good MM work?

 

This is not a simple question, but here is a simple answer: 

 

Personally, I'd prefer to use the "personal rating" as a ranking system for the players, and then use the "ABBA algorithm" for splitting 14 players into 2 teams. Since there are more than 14 players in the wait queue, it is much more complicated, and you can pick tank types etc, but the main idea is quite simple. I'd drop ranking battles and selections for supremacy or not, and same mode and all the selections that further split the wait queue and contribute to longer waiting times. The most important thing is how good the players are, everything else is secondary.  

 

And of course, I'd prefer WG to clearly explain how the MM system works and what it does (not what it does not do!). That's my main (personal) gripe. 

 

Reminders:

 

1. The "personal rating" is a number we all have in their database. Two years ago you could see this number in blitz, then it was banished. But you can still see that number in the WG application "Assistant for World of Tanks Blitz". As an aside, we have no idea what other numbers WG is storing in their database for each one of us, and if any of those numbers are being used by the MM. I am sure they store a lot of other statistics and rankings, but I don't know what exactly they do, so I use the one ranking number I definitely know they have. Don't take this as the only possibility, it is just a starting point.  

 

2. The "ABBA algorithm" is a simple algorithm that is used in playgrounds worldwide. It's a good starting place for having balanced teams for soccer or other playground games, and I know that variations of this are used in computer games as well. Here are some explanations: 

http://bit-player.or...-up-sides-again

 

3. MM is not the simple thing many people believe. A single programmer can create a MM for everything I wrote above in a single day! But of course companies spend much more time programming this, because taking everything into account and producing something that works is much more complicated. It is quite possible that WG is tweaking MM quite often, since it is a server-side algorithm. However, there is always a starting point which is quite simple and yes, you can always give clear explanations about how it works after you finish the implementation. 

 


Edited by CA_vampire, 17 January 2019 - 10:23 PM.


TO_Dominator #54 Posted 17 January 2019 - 11:22 PM

    First Sergeant

  • Players
  • 9154 battles
  • 2,671
  • [YODA_]
  • Member since:
    09-26-2017

View PostCA_vampire, on 17 January 2019 - 10:22 PM, said:

 

This is not a simple question, but here is a simple answer: 

 

Personally, I'd prefer to use the "personal rating" as a ranking system for the players, and then use the "ABBA algorithm" for splitting 14 players into 2 teams. Since there are more than 14 players in the wait queue, it is much more complicated, and you can pick tank types etc, but the main idea is quite simple. I'd drop ranking battles and selections for supremacy or not, and same mode and all the selections that further split the wait queue and contribute to longer waiting times. The most important thing is how good the players are, everything else is secondary.  

 

And of course, I'd prefer WG to clearly explain how the MM system works and what it does (not what it does not do!). That's my main (personal) gripe. 

 

Reminders:

 

1. The "personal rating" is a number we all have in their database. Two years ago you could see this number in blitz, then it was banished. But you can still see that number in the WG application "Assistant for World of Tanks Blitz". As an aside, we have no idea what other numbers WG is storing in their database for each one of us, and if any of those numbers are being used by the MM. I am sure they store a lot of other statistics and rankings, but I don't know what exactly they do, so I use the one ranking number I definitely know they have. Don't take this as the only possibility, it is just a starting point.  

 

2. The "ABBA algorithm" is a simple algorithm that is used in playgrounds worldwide. It's a good starting place for having balanced teams for soccer or other playground games, and I know that variations of this are used in computer games as well. Here are some explanations: 

http://bit-player.or...-up-sides-again

 

3. MM is not the simple thing many people believe. A single programmer can create a MM for everything I wrote above in a single day! But of course companies spend much more time programming this, because taking everything into account and producing something that works is much more complicated. It is quite possible that WG is tweaking MM quite often, since it is a server-side algorithm. However, there is always a starting point which is quite simple and yes, you can always give clear explanations about how it works after you finish the implementation. 

 

 

So you want skill based MM. 

 

That's exactly the "rigged" MM these flat earther are harping about. Wargaming doesn't care about a specific player enough to go after them and ruin their gameplay, especially since that loses them money. So that puts targeting specific players out of the question. It wouldn't be targeting good players either, because then good players wouldn't exist. 

 

Skill based MM is what I consider rigged MM.


"Blitz" can't exist within a world of TDs.

XoBlitzkriegoX #55 Posted 18 January 2019 - 05:01 AM

    Sergeant

  • Players
  • 36607 battles
  • 470
  • [-POD-]
  • Member since:
    11-06-2014

View PostTO_Dominator, on 17 January 2019 - 11:22 PM, said:

 

 

Skill based MM is what I consider rigged MM.

 

 

i bet if your avg dmg and wr were higher, youd be thinking differently.

 


Edited by XoBlitzkriegoX, 18 January 2019 - 05:03 AM.


TO_Dominator #56 Posted 18 January 2019 - 01:36 PM

    First Sergeant

  • Players
  • 9154 battles
  • 2,671
  • [YODA_]
  • Member since:
    09-26-2017

View PostXoBlitzkriegoX, on 18 January 2019 - 05:01 AM, said:

 

 

i bet if your avg dmg and wr were higher, youd be thinking differently.

 

 

No, if my WR were higher, I would be even more convinced that skill based MM was rigged MM, as it will bring all WRs to 50%. 

"Blitz" can't exist within a world of TDs.

Bachus131 #57 Posted 19 January 2019 - 01:20 AM

    Sergeant

  • Players
  • 21763 battles
  • 265
  • [JAWZ]
  • Member since:
    12-28-2017

View PostXoBlitzkriegoX, on 17 January 2019 - 11:01 PM, said:

 

 

i bet if your avg dmg and wr were higher, youd be thinking differently.

 

 

there is only so much damage that can be done in each game so  if you put all high damage players on the same team you would get far less than you do now and Masteries would be much harder to attain. 

 Masteries happen when your team is just good enough not to die in the first two minutes but not so good that they rack up high damage and the enemy are mostly Muppets. 


'Don't follow the lights, they will lead you to your DOOM'  ( Gollum )

RatPATROL_2014 #58 Posted 19 January 2019 - 01:35 AM

    First Sergeant

  • Players
  • 76144 battles
  • 1,060
  • [THUMB]
  • Member since:
    07-01-2014
Yea pathetic makes you not want to battle and no way spend money no no no - I been painting the house so I don’t have time to battle and man it’s bad





Also tagged with matchmaking

1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users