Question about MM and Div's - WTF?

[SolanumTuberosumRex]

This is a game I just played. The single division on my team was formed by two random players in-game, I don't know if they knew each other or not but I think not.

The other team had two 3-man divisions. I've never seen such a disparity before, is this as rare as I think it is?

[Aethervoxx]

24 minutes ago, PsychoClownfish said:

The other team had two 3-man divisions. I've never seen such a disparity before, is this as rare as I think it is?

Not at all. Disparities with divisions in random battles is one of the ways Wedgie can 'arrange' battle results. Your example above is particularly blatant & obvious.

[SodaBubbles]

1 hour ago, PsychoClownfish said:

This is a game I just played. The single division on my team was formed by two random players in-game, I don't know if they knew each other or not but I think not.

The other team had two 3-man divisions. I've never seen such a disparity before, is this as rare as I think it is?

Nope. It's common. It's so common that in 2018 I recorded stats from almost 800 games, and had so many with 3+ divisions I could run simple stats on them. The side with at least a +1 advantage in divisions wins over 60% of the time, I found. Of the major factors influencing match outcome, divs were the most powerful, consistent with the ordinary experience of players.

That is also true of radar: if there are less than 5 radars in total, the side with the +1 advantage wins 60% of the time. Once there are more than 5 radars, the team with the most has a very slight advantage, only visible in stats.

Finally, the old ranked system: teams with at least a +1 advantage in players who had made ranks 1-5 won 60% of the time. I haven't collected stats since they switched to the Sprint mode, I play much less than I used to, and more of it in non-Random modes, Randoms being so awful. 

The fact that WG makes no attempt to balance any of this is WG's way of ensuring that matches are usually frustrating collapses rather than enjoyable, hard-fought matches. I couldn't have imagined a set of developers who detest their own game play so much as these guys. *sigh*

 

 

[Cpt_Andre]

6 hours ago, PsychoClownfish said:

This is a game I just played. The single division on my team was formed by two random players in-game, I don't know if they knew each other or not but I think not.

The other team had two 3-man divisions. I've never seen such a disparity before, is this as rare as I think it is?

You lost against a triple Kurfürst-div, interesting

[DDG_121]

Welcome to the Super MM Komrad.

 

Play ranked. I do that. I'm happy with that.

[principat121]

The rules of the MM allow the numbers of DIV per side to differ by one. 

 

Your team has 1 DIV, the enemy 2 DIV. So, everything is within the communicated rules of the MM. 

 

---

 

When your Top-XP player was mediorce tier9 cruiser and your least XP-scorer a tier11 supercruiser, then you didn't lose this match because of the difference in DIV.

Your team was just plain bad!

Period.

Get over it. 

[DDG_121]

1 hour ago, principat121 said:

When your Top-XP player was mediorce tier9 cruiser and your least XP-scorer a tier11 supercruiser, then you didn't lose this match because of the difference in DIV.

Your team was just plain bad!

Period.

Get over it.

Even if the enemy teams top two are in the same div and 4 of top 6 are from divisions. It is not the division balance. Right komrad?

 

Period.

 

Play ranked to ease the stress.

I hope they don't allow divs in ranked.

 

[principat121]

1 hour ago, Zimbiye said:

Even if the enemy teams top two are in the same div and 4 of top 6 are from divisions. It is not the division balance. Right komrad?

Playing in a DIV amplifies your Skill-niveau. If you are bad, a DIV from bad players is even worse. If you are a good, a DIV from good players is a force to reckon with.

 

And the pure skill was never a parameter for the MM.

Thus, it is in the limits of the MM to match two triple unicum-DIV against a single tomatoe-Duo-DIV.

This match up has a highly predictable outcome, but is nevertheless within the limits of the current MM.

[Yedwy]

10 hours ago, PsychoClownfish said:

This is a game I just played. The single division on my team was formed by two random players in-game, I don't know if they knew each other or not but I think not.

The other team had two 3-man divisions. I've never seen such a disparity before, is this as rare as I think it is?

Nope buisness as usual in wows

[FallenOrchid]

13 minutes ago, principat121 said:

Thus, it is in the limits of the MM to match two triple unicum-DIV against a single tomatoe-Duo-DIV.

This match up has a highly predictable outcome, but is nevertheless within the limits of the current MM.

OP wrote that the div on his side was formed in the match. To my knowledge the total number of divs per side entering the match must not diff by more than one so either also one of the enemy divs was formed in the match or this is a bug of the MM.

[Bainsmit_steel]

Divisions have no place with solo rank que (as they give unfair advantage to division players) but sadly this game will never change that ...or we can play ranked where there are no divisions...

[_Lupastro_]

This is nothing. In a game yesterday I found myself against a division 3x Salem and 2xPetro+Moskva.

 

Obviously no radar for us. Well done MM

[Bear__Necessities]

What is it with players thinking 1 bit of anecdotal evidence suddendly turns their opinions about MM into facts? 

[principat121]

29 minutes ago, FallenOrchid said:

OP wrote that the div on his side was formed in the match. To my knowledge the total number of divs per side entering the match must not diff by more than one so either also one of the enemy divs was formed in the match or this is a bug of the MM.

What do you think: How likely is ot that this could be a bug?

 

A bug so obvious that the forum would be flooded with examples of "numbers of DIV differs by more then one".

[Redcap375]

That is a very poor example. 

 

And im for more balancing when it comes to divs, but that example doesnt say anything your trying to point out. 

 

All it points out is that your team was bad and they prob let 2 kurf steam down a flank. No dds didnt really help and gave more fredom to the kurf. 

 

Its not like you got stomped. I could show you LOTS of one sided matches regarding unicum divs on the same side when they dont even loose a single ship, but this doesnt show any of that. 

[DDG_121]

20 minutes ago, Bear__Necessities said:

What is it with players thinking 1 bit of anecdotal evidence suddendly turns their opinions about MM into facts? 

I've more than 4K matches in this account and my old account even had more.

 

The MM had always been problematic. Even simple solutions cannot be considered. If you log all the matches you play with potatoalert, accepting everyone is playing with open stats, I don't think you'll find any evidence of a good skill based algorithm in there.

 

At least balance radars right? No. Why bother. It is called RANDOM. So random it is.

 

I don't think it is different for Ranked. Maybe WG goes more open with their matchmaking algorithms like they did with container drop rates. But I higly doubt that. Because clearly it involves some mechanics to protect paying players joining with premium ships etc. In some post someone releasd their patent application for MM mechanics. That was in there.

[Bear__Necessities]

1 minute ago, Zimbiye said:

I've more than 4K matches in this account and my old account even had more.

 

The MM had always been problematic. Even simple solutions cannot be considered. If you log all the matches you play with potatoalert, accepting everyone is playing with open stats, I don't think you'll find any evidence of a good skill based algorithm in there.

 

At least balance radars right? No. Why bother. It is called RANDOM. So random it is.

 

I don't think it is different for Ranked. Maybe WG goes more open with their matchmaking algorithms like they did with container drop rates. But I higly doubt that. Because clearly it involves some mechanics to protect paying players joining with premium ships etc. In some post someone releasd their patent application for MM mechanics. That was in there.

A patent does not mean it's used. Just means they are protecting their property. Randoms is really just random. Does not matter over the masses of games if one team occasionally has 2 radar to zero radar. Because something else normally balances it out. Like the other team will have more hydro or stealthier DD's or even the better individual players ect. MM is the best it's ever been. Anyone who's played this game since it started can see that, even if some over the years and masses of threads don't want to admit it.  Ship designs and mechanics are NOT in the realm of MM. MM is not designed to sort what ships might hold what tools or mechanics. It's an arcade shooter and MM is designed to match matches as quick as possible. Lets not ask WG to complicate their algorithms more. We know they can't even release a patch without [edited]over the game. 

[DDG_121]

3 minutes ago, Bear__Necessities said:

Because something else normally balances it out

Not always. If the MM had been a little better, matches with one team just rolling over the other team would be less.

 

IMO they are not that sparse. Good thing I left randoms. Only ranked for me now.

[Bear__Necessities]

2 minutes ago, Zimbiye said:

Not always

Exactly. So it is random. MM will never be perfect. It can't. This game keeps adding in more variables on top of the RNG nature within the game. And Ranked isn't much better. Ranked is just a game of attrition. Not skill. 

[SV_Kompresor]

8 minutes ago, Zimbiye said:

Not always. If the MM had been a little better, matches with one team just rolling over the other team would be less.

How so? Most of the steamrolls happen when teams are relatively balanced wr wise.

[_DemonGuard_]

58 minutes ago, FallenOrchid said:

OP wrote that the div on his side was formed in the match.

And he probably just made this up. For me it looks more like divisions B and C did a sync drop, because they both have players from the same clan and tier.

[FallenOrchid]

1 hour ago, principat121 said:

What do you think: How likely is ot that this could be a bug?

 

A bug so obvious that the forum would be flooded with examples of "numbers of DIV differs by more then one".

Bugs can be triggered often or rarely - and might have slipped into code long ago or just recently.

If the OP is sure the Div on his side was formed in the match and the two on the other side have been formed before the match then he should open a bug.

[FallenOrchid]

1 hour ago, _DemonGuard_ said:

And he probably just made this up. For me it looks more like divisions B and C did a sync drop, because they both have players from the same clan and tier.

Sure this could be. And I cannot remember any match I saw with zero vs. two Divis at start.

[Hugh_Ruka]

4 hours ago, _DemonGuard_ said:

And he probably just made this up. For me it looks more like divisions B and C did a sync drop, because they both have players from the same clan and tier.

Excuse me WHAT ? Since when do you want to form divisions with a tier mismatch ? Also it is most common to see divisions from the same clan.

[Redcap375]

 

 

 

What happens when good Div's are put together. Broken and WG don't give a shite.  Not fussed about Perfect MM, whatever, but for god sake don't put 2 x good divs together SPLIT THEM!