http://mwomercs.com/...stem-explained/
Russ Bullock, on 06 October 2014 - 09:23 AM, said:
I wanted to provide some more explanation on how the new game mode voting system was going to work since I saw lots of incorrect speculation. Many were feeling that certain game modes might completely disappear no matter how they voted, this shouldn't be the case. Please see the explanation below:
1. Starting Bias
The pending game starts with one default bias for all game modes:
Skirmish 1
Assault 1
Conquest 1
2. First group – +4 People - Leader’s Bias: Skirmish, Assault
A group of 4 is added to the game, and the group leaders selection was for Skirmish and Assault. The other group members votes are not transmitted to the matchmaker, so the group leaders selection is counted 4 times:
Skirmish 5 (+4 bias)
Assault 5 (+4 bias)
Conquest 1
3. Second group – +12 people – Leader’s Bias: Conquest
Next, a group of 12 is added to the game, again counting the leader of that groups selections x 12. In this case, the group leader selected only conquest:
Skirmish 5
Assault 5
Conquest 13 (+12 bias)
4. Third group – +8 people – Leader’s Bias: Assault Conquest
Finally a group of 8 is added to complete the game. The leader of this group selected assault and conquest:
Skirmish 5
Assault 13 (+8 bias)
Conquest 21 (+8 bias)
5. Cumulative Distribution Function
So final vote tally is 39 votes. Summing these values up, the following table shows a cumulative distribution function given the users votes.
Skirmish 5 (+0)
Assault 18 (+5)
Conquest 39 (+18)
6. Dice Rolls
The Matchmaker now rolls a random ’39 sided’ die. If the number comes out to 5 or less, the Matchmaker selects Skirmish. If the number is between 6 and 18, the Matchmaker selects Assault. If the number was 19 or greater, then Conquest.
Skirmish ~13% odds
Assault ~34%
Conquest ~53%
In the solo queue - of course each player is added individually instead of using group leaders.
So if you would like to see conquest more frequently, your best course of action might be to vote for Conquest only, that way you increase the likely hood of playing conquest without adding to the chances of the other game modes.
Once this goes live we will be watching the MM closely with the hope that the average ELO difference between group queue matches will reduce noticeably.
Update: Yes a future map voting system would work the same way, so no map should disappear completely but players can influence the frequency.