skill based game joining procedure to insure balanced teams

greenhorse

When you are in ready room you cant join team which has more players than the other team.
This guarantees that player counts in teams will not differ by more than one.

What do you think about this alternative (optional) join procedure:

Total skill count is kept for each team. When a player joins a team his skill number is added
to total team skill count. Player can join only a team which currently has the smallest team
skill count.

#############################################################
For example:

Player with skill 2000 joins marines.
Then next player with skill 500 joins aliens.
Next player (1000) has to join aliens as well.
Next player (600) also goes to aliens.
Now we have 2000 total skill on marine team and 2100 for alien team.
So next player (2000) will go to marines.
After that players (500, 500, 500) go to alien team.
So at this point we have the following teams :

marines : 2000 2000 (total skill 4000)
aliens: 500 1000 600 500 500 500 (total skill 3600)

etc.
##############################################################

Don't know if this will work well. But it will allow for example for smaller higher
skilled teams to play against larger teams consisting of low skilled players.
And it shouldn't allow unbalanced teams.

2cough

Not exactly what you're talking about, but the way day of defeat source does it (I think) is when you late join and random, if skillthe teams are even, the extra person connecting is put on the team with the lowest score. Ns2 keeps track of individual player score, but not the team score per game. I personally liked the way dods did it and wouldn't be opposed to something basic like that.

Would still be nice if maybe when you random, it would take your hive stats into account and put you on the team with the lower accum. skill rating. (Or highest if you suck lol)

coolitic

The problem is, the way skill is calculated in-game does not measure actual skill too well (though it might give some sort of idea)

Nordic

2cough wrote: »

Not exactly what you're talking about, but the way day of defeat source does it (I think) is when you late join and random, if skillthe teams are even, the extra person connecting is put on the team with the lowest score. Ns2 keeps track of individual player score, but not the team score per game. I personally liked the way dods did it and wouldn't be opposed to something basic like that.

Would still be nice if maybe when you random, it would take your hive stats into account and put you on the team with the lower accum. skill rating. (Or highest if you suck lol)

I believe it already does take hive stats into account, but I am not sure. @moultano‌?

moultano

Nordic wrote: »

2cough wrote: »

Not exactly what you're talking about, but the way day of defeat source does it (I think) is when you late join and random, if skillthe teams are even, the extra person connecting is put on the team with the lowest score. Ns2 keeps track of individual player score, but not the team score per game. I personally liked the way dods did it and wouldn't be opposed to something basic like that.

Would still be nice if maybe when you random, it would take your hive stats into account and put you on the team with the lower accum. skill rating. (Or highest if you suck lol)

I believe it already does take hive stats into account, but I am not sure. @moultano‌?

Not yet. I was going to implement that but then I had a baby.

AussieKid

coolitic wrote: »

The problem is, the way skill is calculated in-game does not measure actual skill too well (though it might give some sort of idea)

The problem is that hive and ns2stats are tracking all the details needed for a pretty good ELO system, it simply isn't implemented into the game. If only it took into account a weighting of hours played, win/loss and k/d...

i.e. A player with 10 hours of hive stats-recorded gameplay with 5:2 k/d should not be higher in skill than a player with 1000 hours of hive stats-recorded gameplay with a 2:1 k/d. However surely a player with 300 hours of gameplay and a 5:2 k/d should not be lower in rank than a player with 400 hours of gameplay and a 1:1 k/d.

Now I am well aware k/d is not the deciding factor in how "tactical" a player is in NS2, but it sure does hold the most weight in what the outcome of the game will be.

Hive did address this issue by seemingly incrementing a player's hive score depending on how many matches they play, but do not seem to take into account the player's performance.

Therius

The game doesn't work that way. Four players with a skill of 500 do not equal one player with a skill of 2000. There are many factors at play which require the teams to be even in numbers. No matter how good you are, you cannot be in two places at once.