There are players who suspect that behind a series of losses in League of Legends lies some sort of “matchmaking conspiracy.” One player has now analyzed around 100,000 Master ELO matches to investigate the existence of a “loser queue.”
The lousy tank, the broken mouse, the fluctuating weather – there are plenty of excuses for a crushing defeat in League of Legends. A popular theory suggests that matchmaking is manipulated to prolong losing streaks: the so-called “loser queue.”
The theory states: Developers at Riot want to push the good players who have been around longer. So if you lose a match, you get placed in a team with other losers in the next match.
The “bad” players lose more often. The “good” players enjoy their winning streaks and become more attached to League of Legends. Such systems are called “Engagement Optimized Matchmaking” and there are scientific studies on this (via web.cs.ucla.edu).
Whether such systems are in place is not so easy to detect. Reddit user and astrophysics student “enecotyfanboy” has attempted this by analyzing 100,000 matches and published his findings on Reddit. Here are his conclusions.
In the video, you will find some info about your MMR and how it is generated:
How did the user test? enecotyfanboy collected data from 100,000 matches through an API from Riot. Then he selected 1,000 randomly chosen players and compiled their win/loss history.
He is aware that his methods have certain weaknesses. With the relatively large amount of data, he wanted to smooth out these discrepancies a bit. He also reminds that this is just a Reddit thread and not a scientific study:
What do the results look like? enecotyfanboy concludes that there is no “algorithmically orchestrated win or loss streak mechanism” in LoL.
His results roughly align with the value you would expect from a coin toss based on the players’ win rates (around 55%). There are no statistical anomalies that would indicate a “loser queue.”
However, he also says that he does not have enough data to prove that Riot does not manipulate matchmaking in any way. At least, according to his research, a “loser queue” is less likely.
As a tip, enecotyfanboy suggests: Take more breaks between your games. That way, you will play better.
What do players say about this? The discussions range from serious to funny to fascinating. The mood in the thread is quite good. The post gathered nearly 400 comments.
The most popular comment is, of course, nonsense, as is often the case on Reddit: “No. This analysis means nothing, as the CEO of Riot has classified me and only me personally in the loser queue. Statistically, that’s too irrelevant to be noticed in your sample of 100,000 games.”
By the way, there have been speculations many times about significantly changing the matchmaking in LoL: 4 years ago, a new system was supposed to improve your matchmaking – but frustrated players had it scrapped.
