Skip to content

Riot will you please listen to my suggestion...this will help alot of people in solo q ranked

AchillesofWar··1 votes·775 views

Riot you have a major problem with your solo ranked matchmaking system. their should be no way someone with over 500 games in ranked and in silver or higher should be teamed with players that are either unranked or have less than 10/50/100 games played in ranked depending on how many games you have under your belt. the frustration for both experienced players and the newbies alike is tremendously obstructive to the progress of a seasoned player grinding up the ranks, as well as for the inexperienced players being disheartened by pings/commands being given by his team. This is very detrimental to new players that are trying to step into a world of ranked that is viciously cruel and unforgiving. I'm certain it has shut off many players from continuing to play the game and improve simply for the fear of entering a game with veteran players that are out of their league. This is due to a lack of a filter with experience in ranked for the matchmaking algorithm; something that can easily be fixed! Make this happen Riot, you will be helping both the old and the new summoners enjoy the game even more. -Achillesofwar

9 Comments

FFF12 votes

the only needed metric needed to determine your rank should be your overall nexus killing ability.

You are matched with other players solely based on your overall ability to win. This means that if you are matched with someone who just started to play ranked and he has a win rate that leveled off at 50% at the same rank that your win rate leveled off at 50% then no matter how many games you've played more than him you are at the same capability of winning as that person.

I think the best system is the current one. That is to match you with only people who are just as talented at nexus killing as you are. That means that person who is mechanically better but has poor map awareness or rages when the team is losing early game can most certainly be ranked and matched with people who got the same rank by being really aware of the game state and knowing when to take objectives and such but has a playstyle or mechanical inepitude that requires them to be carried when it comes down to it.

If matchmaking started taking different variables into account to determine who get matched with who then you run into the problem of assigning arbitrary weights to each of the variables and you get something similar to dominion's score stat. It often doesn't come close to reflecting how well you did in the match because it uses variables that weren't optimally weighted. The best determiner of rank should remain asking for each game, the binary question "Did you win? (Y/N)"

ArtoriusIV2 votes

Although I agree with this if this were a single player game(SC2 for example) this team oriented game should look at did YOU win, versus did your team destroy the nexus. How to do that with a program, I do not know, but we have all played games that you felt good about even after losing, and games you felt bad about even though you won. It needs to match players of equal mechanic level. not with equal level of nexus kills since there are too many variables to why you won when there are 10 people involved.

AchillesofWar1 votes

Although I agree with this if this were a single player game(SC2 for example) this team oriented game should look at did YOU win, versus did your team destroy the nexus. How to do that with a program, I do not know, but we have all played games that you felt good about even after losing, and games you felt bad about even though you won. It needs to match players of equal mechanic level. not with equal level of nexus kills since there are too many variables to why you won when there are 10 people involved.

You are correct that individual performance should also be taken into account but like I said before it is too hard to justify being a factor in matchmaking with 100% certainty of a players performance (mechanics).

FFF12 votes

nexus killing is not equivalent to individual performance. Individual performance is but one aspect to nexus killing. see my other post.

ArtoriusIV2 votes

nexus killing is not equivalent to individual performance. Individual performance is but one aspect to nexus killing.

And this is the aspect which matters when matching people together. Do they each do as well as each other on average. for (a bad) example, if you have an average KDA of 1.2, should you be matched with people of average KDA's of 0.2 or 4.3? or should you be matched with people near your individual skill level, whether it results in their nexus dying?

I dislike the idea of using KDA to determine MMR since it will foster people taking less risks, which would have worked if people were willing to take a death for two towers etc...) but it all boils down to, did YOU win/feel good about your performance or did you not/feel poorly about your performance. These parameters you use to determine a good game vs a bad one is what MMR should be based off, not your w/l or KDA.

FFF13 votes

League of Legends is a team game and requires teamwork. It also requires mechanical ability and insight into how the game is progressing. When i say nexus killing I don't mean mechanical skill alone or experience alone or team play alone. nexus killing is a conglomerate skill which in and of itself encompasses everything there is required to be skillful at league of legends.

This is certain because killing the nexus is the final and ultimate objective of any match. Nothing matters to the conclusion of a match other than which nexus dies. All the events leading up to that conclusion and how it occurs is variable and not as relevant to the binary question "Did you win? (Y/N)".

Taking into consideration any other variables into your rank means you are artificially and arbitrarily weighing the metric almost certainly away from the core conglomerate metric that is winning the game. This guarantees any model other than the conglomerate skill will be overall less accurate.

AchillesofWar1 votes

basing the matchmaking off purely nexus killing does not factor in how many times you've lost a game, not because the other team was better, but because they had more experience. there are many mechanically "good" players that are still learning the game and being thrust into silver without having experience is in most cases setting you up for failure. If riot wants to place someone in silver because of their skill or nexus killing thats fine, but place them in matches with other inexperienced silver players so that they can continue to learn the game at their level of play. There are only two factors which riot can extract with absolute certainty to justify a summoners rating. Their wins vs losses (nexus killing) and their experience. KDA ratio should not be used as a factor, cs could possibly be used but its also not 100%

FFF12 votes

their win versus loss rate is dependent on how experienced you are.

N = win rate E =experience C = mechanical skill

N(E,C,...) //win rate is a function of experience, mechanical skill, various other metrics

this means and any one of the independent variables are not solely responsible for the result of the win rate. Win rate already takes every variable into consideration and reading this rate means you don't have to find out how the function N works nor is it useful in order to use N as a metric.

AchillesofWar1 votes

their win versus loss rate is dependent on how experienced you are.

N = win rate E =experience C = mechanical skill

N(E,C,...) //win rate is a function of experience, mechanical skill, various other metrics

this means and any one of the independent variables are not solely responsible for the result of the win rate. Win rate already takes every variable into consideration and reading this rate means you don't have to find out how the function N works nor is it useful in order to use N as a metric.

thats not true because if your getting paired with inexperienced players in silver, your win rate is going to take a hit, if you were always playing with same skill and experience players, then your win rate is an absolute way to measure your mmr...but the way the matchmaking is set up now, N=E+C+O(4) O=teammate with equal C but E is not factored therefore has variance and is the root of the problem with the equation working.

More from Gameplay

More by AchillesofWar