24 replies to this topic

[SqueakyNut]

I have noticed that so many times teams are unfairly matched when it comes to experience.  The program does not take into consideration a players win/loss percentages when creating teams. Although the tanks may be fairly matched, the win/loss is drastically different. This TEAMING of players makes matches very one sided. When a team (as a group)has an average winning percentage of 52% plays a team that has an average winning percentage of 45% the team with the lower percentage always loses... ALWAYS! This is very unfair and in essence the result of the game has been determined prior to the first shot. 

[SpartacusDiablo]

This is a topic that is rarely discussed and should be looked at in more depth.  If only there were a pinned Matchmaking thread where a Wargaming employee explains the Matchmaking process and how it works.  If only...

 

Spoiler

 There is a pinned Matchmaking thread for this discussion.

 

[Posit1ve_]

Not necessarily; I've seen teams with even larger statistical advantages end up getting spanked.

 

Believe it or not; matchmaking is random. Which means over the long run; the only thing that is present in every game is you.

 

I'd say ~80% of games winnable if you play perfectly; so usually; there's something you could've done to win the game. Just keep trying to improve yourself; and you'll find you win more of your matches. Simple. 

 

You can track your own personal progress here: https://www.blitzsta.../com/SqueakyNut

[_chasey]

I had a game the other night with 7 tds on green. So much for the updated mm algorithm

 

[4sfield]

  Everything doesn’t have to be what everone considers fair or what nubs consider fair. That why the game is where it is with its current configuration. 

 

  I don’t have a problem being matched in a game where my team is at a disadvantage by tank type. As far as experience is concerned, it’s every players job to take responsibility and carry their weight in every battle. If they can’t do that then they should drop tiers. If they’re not willing to do that then they deserve to lose.

[VikkoTheTusken]

SqueakyNut, on 11 January 2018 - 02:38 PM, said:

... ALWAYS! 

 

That's a classic thinking trap. It's also not true. Get some mental relaxation by picking up some thinking trap defeating guidance HERE

 

Also, there's an official MM thread HERE

Edited by VikkoTheTusken, 11 January 2018 - 07:57 PM.

[WipWapJaws]

SqueakyNut, on 12 January 2018 - 04:38 AM, said:

I have noticed that so many times teams are unfairly matched when it comes to experience.  The program does not take into consideration a players win/loss percentages when creating teams. Although the tanks may be fairly matched, the win/loss is drastically different. This TEAMING of players makes matches very one sided. When a team (as a group)has an average winning percentage of 52% plays a team that has an average winning percentage of 45% the team with the lower percentage always loses... ALWAYS! This is very unfair and in essence the result of the game has been determined prior to the first shot. 

 

That is an interesting idea, no one has ever thought of it before........

[MRobertsRCS]

Excellent first post.  Welcome to the forums.

[SqueakyNut]

I am hoping an administrator reviews the post. I could not find another link for user feedback. Thanks for your comments. I am a numbers oriented guy and have watched pretty much every game. The numbers don't lie, there are exceptions, but when you see a blow out take a look. Blow outs are almost always mismatched win/loss percentages. 

 

[wolfpack_41]

SpartacusDiablo, on 11 January 2018 - 01:42 PM, said:

This is a topic that is rarely discussed and should be looked at in more depth.  If only there were a pinned Matchmaking thread where a Wargaming employee explains the Matchmaking process and how it works.  If only...

 

Spoiler

 There is a pinned Matchmaking thread for this discussion.

 

 

Yeah bro I just used the wrong one (matchmaking clarification) Check it out a game I tried to play this morning. Think its what you were talking about the wrong way tanks are put in battle. The other team was mostly all pros also. Would like your input. Have a great day.  

[SpartacusDiablo]

wolfpack1941, on 12 January 2018 - 03:05 PM, said:

 

Yeah bro I just used the wrong one (matchmaking clarification) Check it out a game I tried to play this morning. Think its what you were talking about the wrong way tanks are put in battle. The other team was mostly all pros also. Would like your input. Have a great day.  

Okay.  I'll give this a go.  Hypothetically I'd be all for splitting team by win rate.  For example, give each team of seven 1 60%er, 2 50%ers, and 4 40%ers and let the team that comes together best win.  Seems nice but it wouldn't work.  We'd be in que for a match for a small eternity.  Here's why.  If you take 10 players you'll probably a end up with 1 unicum 60%er (and that is a huge maybe), three 50%ers and six sub 50%ers...  Now add 4 more players into the mix.  Based off the numbers above, chances  are you'll have one 50%er and three sub 50%ers.  Now  how do you spread them out equally among the two teams???  But that's not all. Now you need to take into account vehicle tiers, are they Platooning, are the two platoons the same tier, and are the win rates of the platoon members roughly equal.  With all  these limitations the matchmaker simply wouldn't be able to generate matches in anything resembling a timely manner.  We'd be stuck in que forever, people would lose interest and the game would quickly die. Nobody wants that.

[4sfield]

  There is no MM problem. Bad teams because of responsibility issues. Nobody takes ownership of being a bad player, they all just want to assume that playing whatever tier they want is their right and since are no repercussions for it. 

 

  This is the only reason we have bad teams. Guys that aren’t effective play tiers that are out of their wheel house. It flies in the face of the guys that have earned the right to play higher tiers  Then they get downvoted for expressing the obvious by the exact same guys.

 

  Entitlement has ruined more aspects of this game than anything else. The matchmaker is fine, the playerbase is [edited].

[SpartacusDiablo]

4sfield, on 12 January 2018 - 04:27 PM, said:

  There is no MM problem. Bad teams because of responsibility issues. Nobody takes ownership of being a bad player, they all just want to assume that playing whatever tier they want is their right and since are no repercussions for it. 

 

  This is the only reason we have bad teams. Guys that aren’t effective play tiers that are out of their wheel house. It flies in the face of the guys that have earned the right to play higher tiers  Then they get downvoted for expressing the obvious by the exact same guys.

 

  Entitlement has ruined more aspects of this game than anything else. The matchmaker is fine, the playerbase is [edited].

I agree completely with one exception.  Seeing teams that are composed nearly entirely of lights vs heavies or five tank destroyers vs two is a huge problem right now.

 

We definitely have a playerbase problem but we also have a problem with a team / vehicle balance.  The types of tanks should ideally be split evenly per team (or as close as the can get).

Edited by SpartacusDiablo, 12 January 2018 - 09:34 PM.

[SqueakyNut]

Okay.  I'll give this a go.  Hypothetically I'd be all for splitting team by win rate.  For example, give each team of seven 1 60%er, 2 50%ers, and 4 40%ers and let the team that comes together best win.  Seems nice but it wouldn't work.  We'd be in que for a match for a small eternity.  Here's why.  If you take 10 players you'll probably a end up with 1 unicum 60%er (and that is a huge maybe), three 50%ers and six sub 50%ers...  Now add 4 more players into the mix.  Based off the numbers above, chances  are you'll have one 50%er and three sub 50%ers.  Now  how do you spread them out equally among the two teams???  But that's not all. Now you need to take into account vehicle tiers, are they Platooning, are the two platoons the same tier, and are the win rates of the platoon members roughly equal.  With all  these limitations the matchmaker simply wouldn't be able to generate matches in anything resembling a timely manner.  We'd be stuck in que forever, people would lose interest and the game would quickly die. Nobody wants that.

 

 

I can deal with that answer, or any answer that is rational. thank you for the feedback. I am certain that the programers (assuming you are one) are always looking for a way to improve programs. Maybe one day this can be overcome. 

 

 

Edited by SqueakyNut, 13 January 2018 - 03:33 AM.

[SpartacusDiablo]

SqueakyNut, on 12 January 2018 - 10:31 PM, said:

 

 

I can deal with that answer, or any answer that is rational. thank you for the feedback. I am certain that the programers (assuming you are one) are always looking for a way to improve programs. Maybe one day this can be overcome. 

 

 

Just a player. I'm not a programmer.

[RoadKill_177664]

Tank type is not the problem, just call town or hills based on team makeup.

 

Surely they can evenly match better than 5 greens and 2 whites VS 2 greens and 5 whites

 

[____Flying_Dutchman____]

Look i have two ideas that ive been suggesting to people that i think would balance the game and straighten some of the abusive players up. 1st red thumbs enable friendly fire rounds from your team against you. For 1 round on light offenders but follow the abusive ones and a red thumb indicator above their tank in battle so everyone knows they can be removed. Suggestion 2 four man platoons. That would balance play in battle i beleieve..

[celestron2]

consider how MM randomness works over a sequence of games. MM Randomness seems to work out such that about half my games host operators who can work synergistically together.  The other half of game sessions are mercifully quickly forgettable. The ratio is about 50/50, so that’s pretty close to a random selection of user capabilities .  Not bad really. Concentrate on enjoying those games in which some users exercise remarkable situational awareness.  Don’t get too bent about the rest.  

 

And don’t be resistive to moving on to something different for a while. Familiarity can breed contempt.  You may find changes made while you were gone promote a different perspective. The game isn’t what I’m looking for lately either. I’m taking a break. 

Edited by celestron2, 15 January 2018 - 04:25 AM.

[4sfield]

celestron2, on 14 January 2018 - 11:22 PM, said:

consider how MM randomness works over a sequence of games. MM Randomness seems to work out such that about half my games host operators who can work synergistically together.  The other half of game sessions are mercifully quickly forgettable. The ratio is about 50/50, so that’s pretty close to a random selection of user capabilities .  Not bad really. Concentrate on enjoying those games in which some users exercise remarkable situational awareness.  Don’t get too bent about the rest.  

 

And don’t be resistive to moving on to something different for a while. Familiarity can breed contempt.  You may find changes made while you were gone promote a different perspective. The game isn’t what I’m looking for lately either. I’m taking a break. 

 

  So, the playerbase sux and you shouldn’t get honked off about it?

 

  Nice revelation Einstein, nobody has ever come up with that before.

[Gregarioun]

SpartacusDiablo, on 12 January 2018 - 08:57 PM, said:

Okay.  I'll give this a go.  Hypothetically I'd be all for splitting team by win rate.  For example, give each team of seven 1 60%er, 2 50%ers, and 4 40%ers and let the team that comes together best win.  Seems nice but it wouldn't work.  We'd be in que for a match for a small eternity.  Here's why.  If you take 10 players you'll probably a end up with 1 unicum 60%er (and that is a huge maybe), three 50%ers and six sub 50%ers...  Now add 4 more players into the mix.  Based off the numbers above, chances  are you'll have one 50%er and three sub 50%ers.  Now  how do you spread them out equally among the two teams???  But that's not all. Now you need to take into account vehicle tiers, are they Platooning, are the two platoons the same tier, and are the win rates of the platoon members roughly equal.  With all  these limitations the matchmaker simply wouldn't be able to generate matches in anything resembling a timely manner.  We'd be stuck in que forever, people would lose interest and the game would quickly die. Nobody wants that.

This is completely untrue.  There is a very simple way to get around this:  you split the teams based on winrate and battles AFTER the initial matchmaking which is in place now.  Because the tanks are already balanced, you simply switch one player for the other to get as close to a 'numerical' balance as possible.  That way you avoid seeing 4 >%55 players on one team with the other having negatives.  I don't think the people over at Wargaming thought this through very well.