When Elo includes matches played as a major factor maybe the most important one, it should result in better matches.
Right now it seems to be as if tonnage, Kill/death ratio and/or Win/loss ratio are the most important factors in matchmaking, that leads to PUGdrops where players with over 1000 matches are playing with and against players with 10 matches but the same ratios.
This often results in extreme games (12-1 kills etc.)
If you add the amount of times played as the most important factor followed by the ratios you should get a more even matchmaking, sure there will be matches were someone with a 0.2 kill/death ratio but 900 games is paired with ppl with 1.9 ratio, but this is preferable, because players with more games should be able to play better and help more, maybe the guys with the bad ratios are awesome light pilots or support players who just don't get that many kills.
What are your thoughts?
1
Change Elo That It Includes Matches Played As A Major Factor
Started by ChaosAptom, Jan 05 2014 05:32 AM
3 replies to this topic
#1
Posted 05 January 2014 - 05:32 AM
#2
Posted 05 January 2014 - 05:52 AM
Well, before making suggestions like these, you should wander over to the command chair section and read up on how our Elo implementation works, as you clearly have no clue. I don't mean that as an insult, just that how you think it works is so far from how it works that your suggestion makes no sense.
#3
Posted 05 January 2014 - 03:29 PM
ELO = Win/Loss raito per weight class (start @38min)
http://soundcloud.co...-heat-scale-elo
hope this helps.
http://soundcloud.co...-heat-scale-elo
hope this helps.
Edited by Warblood, 05 January 2014 - 03:30 PM.
#4
Posted 12 January 2014 - 03:07 AM
read up on Elo and my point stands, the matchmaker does not include games played (or not as much as it should) and this results in teaming up with beginners.
I can see the point in doing it the way it works now(mixing up the player base).
I would also like to see how it work out if matches played is added as a more important factor, mind you not the only factor, as win/loss per weight seems ok.
I can see the point in doing it the way it works now(mixing up the player base).
I would also like to see how it work out if matches played is added as a more important factor, mind you not the only factor, as win/loss per weight seems ok.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users