Hi,
Problem:
ELO is supposed to reflect the ability of the player when teams are being created. ELO is adjusted up and down based on win/loss. Win/loss is dependent on individual ability AND team play. Team play is greatly aided by being grouped with voice communications. Many people play BOTH group and solo games ... as a result their ELO number does not reflect the player's actual ability in the game and thus is not a good measure of balance for the people who play both group and solo ... it makes match making more difficult.
Suggestion:
Introduce "Effective ELO" to the match making process. The "Effective ELO for a player is their ELO multiplied by a grouping factor. The match-maker knows which players are grouped. It knows the size of the groups. It knows the player's base ELO. This base ELO can be multiplied by a factor to determine effective ELO. This factor should go up with group size.
e.g.
2 person group: Effective ELO = 1.1 x ELO (for the grouped people)
3 person group: Effective ELO = 1.3 x ELO
4 person group: Effective ELO = 1.5 x ELO
The 1.1, 1.3 and 1.5 values listed above need to be determined from match heuristics and statistics. Or maybe 1.1, 1.2 and 1.3 would be better values to start with.
The goal of the ELO system is to match up two more or less evenly balanced teams but groups skew this result if the ELO used to calculate match balance does not include an enhancement factor for grouping.
After the match, the players base ELO would be adjusted as usual.
The objective of this suggestion is to try to generate ELO values for all players that are more representative of their individual playing abilities which will allow the match maker to use the base ELO values as a measure for both solo and grouped play.


[Suggesion] Include The Effect Of Groups In Elo Match Making ...
Started by Mawai, Mar 19 2013 06:08 AM
No replies to this topic
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users