Id love to hear from riot if they have any accounting for experience in their MM system. That is to say if the # of games played influences if you get put VS players with a similar # of games played. I feel this is something that reduces the lopsidedness of matchmakers for a team game like leauge.
Also something I never see talked about enough. How Duoing fucks the system up a bit. How does it balance 2 diamonds queing up with a Bronze hommie in normals? Duoing in ranked, it really should not be off tiers but the hidden MMR of accounts. You can be gold 1 with plat MMR and que with a Silver hommie with bronze MMR? where's that gonna put you guys mid silver low gold? I don't like it, and im a guy who ques with hommies all the time cuz I do enjoy playing with friends even when they are awfull. Look there is a reason riots pushing for clash for team play, its easier to balance MMR for the team as a whole then a mishmash of random players, its like their old Ranked teams with some modifications from what they learned from running flex que, with a carrot to encourage folks via free stuff. Going forward Id like harsher Duo requirements build off MMR or I would LOVE its complete removal from solo que. Its fine in flex.
On the flip side, man, so many players have no idea about how elo systems work. it needs games played to get the data to judge folks. The matchmaker has no clue what a players skill is at the start and it takes games for it to start getting accurate in its placement of any one player. Inherent to the system its gonna be lopsided matches, but shit its def better then my wc3 custom game days where there's pure chaos in matches. There's a good reason everybody downloaded 3rd party programs for elo when they where serious about hosting and playing Dota allstar games. You think Riots MM is one sided? Try being a "bronze" player solo vs a 5 man premade "diamond" level clan with your teammates being literally any random MMR. End of the day even a basic systems gonna be superior to none.
Now if we could just get people to understand how unsympathetic the system is and that it does not proactively look to rig matches but just doesn't care if any individual games well balanced because that's just part of how it gets better at making the matches less unbalanced for each player participating in that match. as the game goes on, Its self correcting off large numbers, Bad players go down, good ones rise. There are issues that may or may not have been addressed by riot id love to know more about the system, but end of the day if its just a basic team elo system you should assume that it does not have players accurately placed until they have a few hundred games in any que and yet this is still superior to not attempts to build fair teams.
Oh and we dealing with hundred+ games. a streak is not 3-4 F**King games for the love of all that is good. Pretty please y'all try to set a expectation relevant to reality. 3 matches is not significant, your talking like what, 1 promo or 60 lp? That's nothing. its just a insignificant number of matches your proud of or upset with. Take a look at coin flips, do 100 at 50% and your going to see it filled with 3+ "streaks" of heads and tails. Its to low of a # to put value on when discussing climbing the ladder. Please understand this and stop whining about this pattern of win streaks and loss streaks your seeing, they are not a dam streak unless you play so pitifully few games of ranked that youd be pretty much incapable of climbing the ladder even if most your matches are wins. 30 games, 20 wins 10 losses? 66% win rate yet your going up what 200 lp? so... like... 2 divisions-ish? congrats if you started in silver 3 after placements your now silver 2 or gold 5, with your hidden MMR likely being silver 1 regardless. My goodness please dont complain about streaks and rigged matchmaker more when you don't have the games played to even climb at what most of us would consider a pretty good win %.
