Seriously wargaming, how hard is it to balance MM???

[Just_the_tip_16]

I think a middle school computer programming class could figure out that 3 Yamatos vs. 3 Tirpitz's is not quite "balanced."   2 minutes into the game I lose a turret to a blind salvo into our spawn point, and in under 10 minutes into the game I had ZERO turrets left to fire.  That's right folks...all 4 turrets on my Tirpitz are out of action.  Not incapacitated, but blown up to where they won't fire anymore.  So now I'm in anarmored but very slow and not maneuverable destroyer hoping to spam a few torps into the enemy team.   Not only was MM jacked up, but the typical lopsided RNGesus was also jacked up that game.

 

But working as intended....

 

 

Its really a joke anymore at Wargaming's strategy.  Fast lopsided games to cater to the speed grinder crowd that doesn't play to win, just plays to get credits and XP to unlock the next ship.

[Vinidoanker]

Ask your fellow tirpits friend. to use Yamato..there, 3 vs 3 Yamato potato..Its your freaking choice when u click random battle.expect anything..dude my yamato vs 3 shimikaze..hows that sound? move on like other told me..

[gurudennis]

Oh, there's any number of paid stooges on this forum who will tell you it's like rocket surgery.

[Deathskyz]

Oh, there's any number of paid stooges on this forum who will tell you it's like rocket surgery.

 

But if it isnt why arent there so many games just like this if it was all so easy?

I mean if we follow true to the OP that any middle school programming class could easily program that.

Where are they? I mean if they made a game that catered to people's wishes so easily they could earn a Trumpload of Money.

Who doesnt want a Trumpload of Money?

[Macabe]

You could completely mirror teams and it wouldn't be balanced.

[IceSerpen7]

Oh, there's any number of paid stooges on this forum who will tell you it's like rocket surgery.

 

The "rocket surgery" part is not about making a balanced MM, but about dealing with all ADHD kids raging on the forums about queue times.

[CommodoreKang]

It's not really rocket surgery, it's just that the current MM emphasizes short wait times for anyone, playing any class.  Class and tier restrictions are entirely feasible, it's just the tradeoff is time spent waiting. 

 

 

[Sir_Godz]

It's not really rocket surgery, it's just that the current MM emphasizes short wait times for anyone, playing any class.  Class and tier restrictions are entirely feasible, it's just the tradeoff is time spent waiting. 

 

 

 

extra seconds only that at least give you a chance of a decent match.

[TenguBlade]

I think a middle school computer programming class could figure out that 3 Yamatos vs. 3 Tirpitz's is not quite "balanced."   2 minutes into the game I lose a turret to a blind salvo into our spawn point, and in under 10 minutes into the game I had ZERO turrets left to fire.  That's right folks...all 4 turrets on my Tirpitz are out of action.  Not incapacitated, but blown up to where they won't fire anymore.  So now I'm in anarmored but very slow and not maneuverable destroyer hoping to spam a few torps into the enemy team.   Not only was MM jacked up, but the typical lopsided RNGesus was also jacked up that game.

Have you tried coding a program that sorts anything whatsoever by any qualifications?  Go code a bubble sort or something and come back and tell us that it's so easy; that's what I did in middle school, and it's FAR less complicated than any matchmaking algorithm.

 

This might have more to do with WG not telling us everything that matchmaking is based on than any fallacy on the algorithm's part.

[atPrick__]

 

 

[VGLance]

 

The "rocket surgery" part is not about making a balanced MM, but about dealing with all ADHD kids raging on the forums about queue times.

 

IceSerpen7 keeps bringing this up because debate after debate, he still lacks the cognitive ability to comprehend that queue times would be completely unaffected since the MM adjustment would occur AFTER the 24 players have been picked.  It's not about creating the perfect 50/50 match.  It's about eliminating the outrageous, completely unnecessary 3 yamatos vs. 3 tirpitz scenarios as the OP described.  If they are not in divisions, there is absolutely no reason not to trade one for the other so you have 2 yamatos and 1 tirpitz vs 1 yamato and 2 tirpitz.  Again, not 50/50 perfect, but a HUGE step in normalization.  My MM algorithm would further move toward balance by making it so that it would take the strongest of the three yamato players by efficiency rating and move him over with the 2 tirpitz players.

 

All of which would happen in a blink of an eye.

 

IceSerpen7's tunnel vision and refusal to let go of his idea that balancing can only occur prior to the 24 ships being picked is pretty embarrassing.  Others have tried to explain both the concept and how sad he's making himself look by not grasping it and he still fights so damned hard to make his case.

 

And yes, my algorithm can be written by a high school programming hobbyist.  And I know because I can program it myself in C++.  This is strictly a refusal on WG's part.  And you can tell because they never engage in an open dialogue/forum about it with the community no matter how many times over the years both in WoWs and WoTs people have been screaming for them to do so.  It's one thing if they refuse to do it, it's a blatant demonstration of a massive customer relations blind spot to not simply talk about such a major component of the game.

[Tekina_]

Have you tried coding a program that sorts anything whatsoever by any qualifications?  Go code a bubble search or something and come back and tell us that it's so easy; that's what I did in middle school, and it's FAR less complicated than any matchmaking algorithm.

 

This might have more to do with WG not telling us everything that matchmaking is based on than any fallacy on the algorithm's part.

 

A bubble search is not a thing. You are thinking of a bubble sort. Learn your terminology before you try to preach. 

 

The biggest issues with MMing are the small pool of players vs the time MMing will wait as well as how it handles divisions in general, not the algorithm. Generally, given a reasonable pool of players and no divs MMing does a decent enough job. Make high tier matches fun so more people play them and get rid of divisions(or heavily revise them) and MMing should work fine.

[gurudennis]

Have you tried coding a program that sorts anything whatsoever by any qualifications?  Go code a bubble search or something and come back and tell us that it's so easy; that's what I did in middle school, and it's FAR less complicated than any matchmaking algorithm.

 

This might have more to do with WG not telling us everything that matchmaking is based on than any fallacy on the algorithm's part.

 

I have. Software developer reporting in 

 

The MM is not trivial but simple enough and well-known enough of an algorithmic problem that it should be in a better state by now considering how WG aren't exactly a bunch of amateurs. Either it's not a priority or it's working as intended.

 

Edit: just noticed "bubble search". GG.

[giovybez]

Slak did you guys win?

[CommodoreKang]

 

And yes, my algorithm can be written by a high school programming hobbyist.  And I know because I can program it myself in C++.  This is strictly a refusal on WG's part.  And you can tell because they never engage in an open dialogue/forum about it with the community no matter how many times over the years both in WoWs and WoTs people have been screaming for them to do so.  It's one thing if they refuse to do it, it's a blatant demonstration of a massive customer relations blind spot to not simply talk about such a major component of the game.

 

It's not about the technical difficulty or coding wizardry.  It's about trade offs of various sorts.  Which are pretty obvious if you think beyond the "i want" mentality.

[gurudennis]

I wonder how many more people with a basic 1000-feet understanding of software will comment about its complexity because it seems so complex to them.

[atPrick__]

Divisions screw up the MM, thats the biggest issue. 

 

Otherwise I enjoy some imbalance, what fun is the game if it always even.

Slak did you guys win?

 

we did, we complained about the crap MM and then rushed the cap and stayed away from the slow tomatos.

[Takamachi_Nanoha_sg]

So many people complain about bad MM throw them into a suck team which can't win at all, but where are those people on the other side of the MM? You should have equal chance of being put into the favored team as well. So those who enjoyed slaughtering the enemy team will just keep their mouth shut and only start to complain when it affects their own self?

If you get roughly equal number of good and bad team matchup, I view it as a fine thing. Its not a guaranteed win or lose, I have plenty of screenshots to show how we won in a failed MM...

Well OP if you think a middle school programming lesson can do it better, please by all means send a job application letter and your resume to WG and I'm sure you will be well paid. Or if you are not yet the age to take that lesson, recommend your seniors to WG, they will pay you referral fees.

For Tirpitz...I never afraid of fighting a Yammy in my Tirpitz though...

[VGLance]

 

It's not about the technical difficulty or coding wizardry.  It's about trade offs of various sorts.  Which are pretty obvious if you think beyond the "i want" mentality.

 

Now all you need to do to complete the argument is to state your thesis and support it with factual evidence.  Otherwise, it's just rhetoric.

[gurudennis]

So many people complain about bad MM throw them into a suck team which can't win at all, but where are those people on the other side of the MM? You should have equal chance of being put into the favored team as well. So those who enjoyed slaughtering the enemy team will just keep their mouth shut and only start to complain when it affects their own self?

If you get roughly equal number of good and bad team matchup, I view it as a fine thing. Its not a guaranteed win or lose, I have plenty of screenshots to show how we won in a failed MM...

Well OP if you think a middle school programming lesson can do it better, please by all means send a job application letter and your resume to WG and I'm sure you will be well paid. Or if you are not yet the age to take that lesson, recommend your seniors to WG, they will pay you referral fees.

For Tirpitz...I never afraid of fighting a Yammy in my Tirpitz though...

 

I'll explain with an extreme example. Half the time the MM puts you against a Yamato in your Erie, and half the time you are in your Yamato against an enemy Erie. You win 50% of the time, so it must be OK, right?

[CommodoreKang]

 

Now all you need to do to complete the argument is to state your thesis and support it with factual evidence.  Otherwise, it's just rhetoric.

 

On one end, you have the current MM.  Basic philosophy is get as many players into matches as quickly as possible.  If we pick mid tier as an example, I can't recall the last time I had to wait more than a minute or so.  Usually, it's less than a minute.

 

On the other end, you have a completely locked down, you must have x ratio of n tier of these classes to make a match.  At peak times and middle tiers, you could probably do that and not have a huge impact but there would be times when certain people in certain tiers would have to wait an unknown amount of time.  At slow times and in high tiers, you might always have to wait a good while for a match.

 

Obviously, there is a middle ground.  But any deviation away from the current, wide open MM, is likely to create the potential for someone to have to wait more than they'd like.  If it makes a significant difference in match outcomes, maybe that would be worth it.  My experience is you're either on the fail team or you're not.  And, until it becomes apparent to me that team comp is a major factor in wins, I'm OK with the "everybody gets in ASAP" approach.

 

[gurudennis]

Considering how aggravating the MM can be, and how much anxiety pressing Battle gives me personally because of it, I'd rather wait for 2-3 minutes than end up in a crapshoot every few matches as I do now. Wouldn't you? Isn't it obvious?

[CommodoreKang]

Considering how aggravating the MM can be, and how much anxiety pressing Battle gives me personally because of it, I'd rather wait for 2-3 minutes than end up in a crapshoot every few matches as I do now. Wouldn't you? Isn't it obvious?

 

What if you waited 2 or 3 minutes and still ended up in a crap shoot?

[gurudennis]

Then I'd wait 2 or 3 minutes longer. Except that at the moment the median wait for me is under a minute. There's definitely some wiggle room.

[Takamachi_Nanoha_sg]

 

I'll explain with an extreme example. Half the time the MM puts you against a Yamato in your Erie, and half the time you are in your Yamato against an enemy Erie. You win 50% of the time, so it must be OK, right?

The logic is wrong. It's should be you in he same ship, say the Tirpitz. Give a sample size of 900 games for example, your Tirpitz faces tier 9 and 10 ships like Yamato and Zao etc in about 305 games; and meet same tier Amagi, New Orleans etc. in about 293 games; then the rest 302 games you are fighting with tier 6 and 7s like New Orleans, Myoko etc..this is a fine arrangement.

it does not apply to different ships in this logic.