How does the team autobalance work?

[Lennu]

Is it just about the average level of players in the team or K/D, or maybe both?

[Elindor]

Re: How does the team autobalance work?

It doesn't.

Biggest negative to cRPG at the moment.  Devs need to fix this or at least get it right in the next game.

Some rounds I will literally be ASTOUNDED at what it has decided to do with the teams in an effort to "balance them".
This is worse on siege because it rebalances (or unbalances) each round.

[CrazyCracka420]

Haha came to say "it doesn't", see that's already been covered.

I don't know what the formula is exactly, but here's my take:

It takes the first round heavily into account and will try to move people with relatively even K:D's to balance each other out.  It will try to put people with the same banner on the same team regardless of any other factors.  The formula's idea of balance is when one team is raping the other, to move a bunch of players to the losing team so they now rape the team that was winning the first couple rounds (so it way overcompensates).

To me the formula could be improved and still be pretty simple:

Items to the top of the list are weighted more heavily (with the exception of the banner balance, it would always try to keep people with the same banner on the same team whenever possible, but not at the expense of balanced teams):

K:D of player over the last 3 maps
Class of player (based on items equipped)
Level of character

[Elindor]

I'm honestly now sure what it does now....

But this is how I could see it working.

List of weighted factors to determine balancing (items at top of list have more weight than those below):

1. Banner balance
--- Evaluate result of banner balance and remaining players ---
2. Current score
3. Current KDR

Since banner balance is the strongest factor in the equation, what it needs to do is sort that part out FIRST and then evaluate the result - if the banner balance part of the equation leaves one team VERY STRONG in terms of score/kdr/numbers, then it needs to throw any other players who are doing well in score/kdr on the OTHER TEAM

This is the biggest current problem with the balancer and its usage of banner balance.  It will throw CLAN X all on one team and they are mostly all doing very well, but it doesn't seem to properly counter that by giving the other team enough "powerful" players to counter CLAN X.   If CLAN X is dominating it's members will typically naturally end up with good score/kdr.  Between that and numbers, the balancer needs to counter their effect by giving the other team as many players who are performing well as it can.  Surprisingly though, it often fails to do this and will put CLAN X -AND- many other well performing players (not in Clan X) all on one team, and barely any sufficient counters on the opposite team.

An alternative to using current score/kdr would be for the servers to track each player over time (kinda like battle tracks your KDR) - except no one, even you, would be able to see this number.  Basically it would be average score per map and average KDR (one for battle and one for siege).

Then, the equation would use THIS information in the equation instead of CURRENT score/kdr.

NOTE : I would actually be ok with these additional statistics (average score per round and/or map, and siege KDR being visible to the player on their character page like their battle KDR currently is.  I would not make it public just like the battle one isnt just because that leads to a lot of playing for the wrong reasons).

[Tojo]

An alternative to using current score/kdr would be for the servers to track each player over time (kinda like battle tracks your KDR) - except no one, even you, would be able to see this number.  Basically it would be average score per map and average KDR (one for battle and one for siege).

Then, the equation would use THIS information in the equation instead of CURRENT score/kdr.

NOTE : I would actually be ok with these additional statistics (average score per round and/or map, and siege KDR being visible to the player on their character page like their battle KDR currently is.  I would not make it public just like the battle one isnt just because that leads to a lot of playing for the wrong reasons).

^ This...
I want to hear about Elindor's 99,000/100 KDR  on siege and how he isnt getting any points on the scoreboard.

Also I would like to see a fairer numbers on teams. I have noticed if the server gives one team a bunch of good players, it will give the opposite team MORE players. Sure, it's nice to pad your KD but fighting legions of peasants seems unfair.

[Elindor]

^ This...
I want to hear about Elindor's 99,000/100 KDR  on siege and how he isnt getting any points on the scoreboard.

 

Complaining about lack of points is the catalyst that propels me into getting valor so much...you should try it.

[Vibe]

it's goal is to make the game very frustrating for one team

[Smoothrich]

It goes by your character's KD on the webpage more than anything as far as I can tell, along with your level somewhat

This is why good players always go on the shit team and automatically lose unless they are getting valor or banner stacking, making the game pretty bad for solo play

[Elindor]

I'm glad that we're all in agreement that it does one of the following:

A) Doesn't work

- and/or -

B) Works poorly enough to merit a serious reworking

[Rico]

Class of player (based on items equipped)
Level of character

=o there are no "my old friendcher" teams?

[Sir_Hans]

I was playing a match on NA1

And after the first wave of "balancing" after the first round...

One team had a combined kill score of -1 (one person on the team had one kill, two people had -1 from TK's.), While the other team had a combined kill score of about 20.

Needless to say, the team with -1 combined kill score didn't win any rounds for the entire map.

Calling it "balance" is a crude joke. The game would be better off if there was no banner balance/autobalance whatsoever.

[Elindor]

Yeah Hans I've seen similar things.

On NA2 you will often see something like this :

TEAM 1

PLAYER 1 = 200 pts / 20 and 8 (Clan x)
PLAYER 2 = 180 pts / 15 and 7 (Clan x)
PLAYER 3 = 165 pts / 12 and 6
PLAYER 4 = 125 pts / 10 and 2 (Clan x)
PLAYER 5 = 103 pts / 5 and 2
etc...

balanced vs something like this :

TEAM 2

PLAYER 1 = 105 pts / 12 and 6
PLAYER 2 = 76 pts   / 8 and 6
PLAYER 3 = 55 pts   / 3 and 5
PLAYER 4 = 35 pts   / 4 and 6
PLAYER 5 = 34 pts   / 3 and 1
etc...

----------------------------------------------------------------------

I HAVE AN IDEA
Let's all start screenshoting when we see these monstrosities of balancing.  Even if it doesn't effect any change, it will still be an interesting thing to see....

[Shaksie]

I cry with laughter every time it completely switches teams; swaps team a to team b and thinks that it's balanced.
This and the gold/xp gain system are probably the worst things about this mod.

[Lennu]

I was just wondering that should the equipment cost have an effect aswell? I mean that most clans are stacking fully loomed heavy armor, and that way influencing the server instead just being a skilled group working together. Randomers will be split quite evenly to both teams, but 7 players using the same banner, with 70+ armor will twist the balance.

[CrazyCracka420]

I cry with laughter every time it completely switches teams; swaps team a to team b and thinks that it's balanced.
This and the gold/xp gain system are probably the worst things about this mod.

According to the formula it is balanced.  The winning team moves to the losing team and now the losing team starts winning (evens out the total wins for the teams to be balanced)