Idea for MMR system that rewards players for playing well instead of using pure win/loss ratio

The Real OnlyGP··3 votes·805 views

New players entering ranked for the first time must have at least 5 games played in each role in draft pick before being able to queue for placements. This ensures all players have familiarity with each role so they do not troll if they get auto filled because "they dont know how to jungle". Before entering placements, players will select their 3 best performing roles and 3 top performing champions for each role to calculate baseline MMR. Baseline MMR is calculated based on their win/loss ratio, KDA, and average champion rating. Obviously an average of all 9 champions would have to be calculated but this ensures players have big enough champion pool for each role so they cant say they cant play a role since their champ is banned. Once placed, LP gains and loses are calculated using win/loss, KDA, and champion rating for each match. This way players who perform extremely well in a match are rewarded with a high LP gain and their teammate who did nothing and was simply carried receives a much smaller LP gain. LP loss will work the same way, in the sense that a player who doesnt know how to play safe and feeds will lose more LP than someone who performs extremely well but cannot carry 4 other players. I understand that using KDA to determine MMR is risky since it will give players incentive to farm kills instead of take objectives, but hopefully the champion rating and win/loss would keep it in balance. Also a system in which players in promos are only matched against other players in promos would ensure the most competitive experience since players wont have to worry about losing series due to trolls. It may not be the perfect system, but a system that rewards individual players for performing to the best of their abilities instead of a system entirely dependent on how sane their teammates are is all I ask.

7 Comments

Canonic3 votes

This is not a new concept. The problem is, there is no way to measure all if the things that a person could be doing to help their team win, that don't " count" for anything. K/d/a isn't necessarily indicative of their performance.

Karunamon1 votes

Isn't that basically the whole point of the mastery and grade system?

DestructoDave2 votes

If this game rewarded me for ONLY playing well, I would be the biggest KDA player you could imagine and never do anything risky to help my team win. Thats why systems based on individual performance will never work out. Id get my 7/0/10 and sit in a sidelane afk farming for 35 minutes unless my team was pushing to win.

rtbf704898321 votes

I agree with this post in most of the determining factors of player performance mitigating win/loss of LP. Although I think dmg to Champions compared to average dmg to champions in the role would be better than kda. The current system punishes players heavily for the factors they cannot control.

Sona Ping2 votes

[{quoted}](name=Notion Z,realm=NA,application-id=3ErqAdtq,discussion-id=rZrPcaRO,comment-id=0001,timestamp=2018-04-20T19:20:11.766+0000)

I agree with this post in most of the determining factors of player performance mitigating win/loss of LP. Although I think dmg to Champions compared to average dmg to champions in the role would be better than kda. The current system punishes players heavily for the factors they cannot control.

And instead you would have the system heavily punish players for a different factor they can control: Their specific champion choice. Not their performance. Those are both terrible ideas because neither one of those match playing correctly and making the correct decisions to win and worse yet, incentivize doing the opposite.

Metal Janna1 votes

The first part of your post seems to be predicated on the notion that all players should be able to play more than one or two roles in order to avoid awkward situations caused by autofill. But can't Mid, Top, and ADC mains swap lanes with little difficulty? If so, then the only problematic positions to be filled into are support and jungler. So make the matchmaking algorithm (for ranked) toss out any matches that have one team with nobody who has played fewer than 3 games as the jungler. That way there's always a possible lane swap that can fill the jg slot if someone with no experience gets filled into it. As for support, everyone should know support as a secondary, so if someone gets filled support and nobody on the team knows how to support, then that team deserves to lose.

DestructoDave1 votes

{quoted}

The first part of your post seems to be predicated on the notion that all players should be able to play more than one or two roles in order to avoid awkward situations caused by autofill. But can't Mid, Top, and ADC mains swap lanes with little difficulty? If so, then the only problematic positions to be filled into are support and jungler. So make the matchmaking algorithm (for ranked) toss out any matches that have one team with nobody who has played fewer than 3 games as the jungler. That way there's always a possible lane swap that can fill the jg slot if someone with no experience gets filled into it. As for support, everyone should know support as a secondary, so if someone gets filled support and nobody on the team knows how to support, then that team deserves to lose.

Support isnt even hard. I honestly believe people sabotage themselves on purpose when they are filled support, just to validate to themselves they shouldnt be supporting. "See, I told people I shouldn't be supporting!" Its a crock. You are literally ONLY focusing on half the lane. You harass and thats it. Its the equivalent of going mid and not csing; just throwing your spells at the enemy mid laner. Its ARAM basically.

The only role in the game that is actually different is Jungle. Every lane role is nearly the same. There is some things here and there that are diff, but I'll be damned if you can lane ANYWHERE and not be able to support. Jungle is the real role that requires a different skillset.

More from Gameplay