There’s a whoooole wall of text explaining the fundamentals of rating systems and matchmaking systems, as well as a report covering other basics and more detailed jargon.
Right, so I figured I’ll write another wall of text (for real this time), by explaining rating systems and the fallacy of presuming only more people can speed up queue times.
History
Games tend to use the chess rating systems as a basis for their own adaptions. What a rating system is, is a mathematical formula, with the philosophical purpose of attaching a number that represents a person’s skill, to the person.
This means that everyone has a rating that represents a person’s hypothetical “tr…
https://www.microsoft.com/en-us/research/wp-content/uploads/2006/10/Game-Developer-Feature-Article-Graepel-Herbrich.pdf
You don’t need to try to obfuscate it with your platitudes and meaningless speculations.
Check out his reports too. (i.e. Mark Glickman’s)
Onigari:
Ok I’ll do you a favor and streamline it.
With the constant manipulation of metrics and rating ranges, admins of a rating system can inflate or deflate the rating pool at will.
With the use of rating ranges in matchmaking in this game in particular, and how it starts off limited to a predefined range when you first queue and then expands the search parameters the longer you stay in the queue, it means the amount of these ranges in total in the rating system and how long it takes for the search parameters to expand decides how quick the queue is.
Regardless of how many people are queuing.
For example, let’s say you have 20 different rating ranges. But due to the average wait time taking precedence, and the queue isn’t optimized for the player pool, the wait time then increases. Cut that down in half to 10 rating ranges, it means it’ll go faster.
So long queue times is because the system isn’t calibrated as it should be for the player pool size.
The only way more players helps this, is if the system is calibrated for a larger player pool size. But that doesn’t mean more players = faster queues, it means the calibration of it is what decides the speed.
The more rating ranges in the matchmaking (no, that does not mean the ranges for the titles, it means the ranges used for matchmaking which tend to be much smaller in sizes), the more average the skill level of each matchup will be. In other words, it’ll become more “fair” in terms of chance to win. But it needs more and more players, because each added rating range in the matchmaking takes longer time in the matchmaking process.
While fewer rating ranges in the matchmaking process, it means there’ll be a larger difference of skill and a more unbalanced chance to win in the matchups, but it’ll go faster in the matchmaking process which is better for fewer players.
Onigari:
(The accessibility of hosting services also affects the matchmaking speed btw, but that kinda goes without saying. Can’t matchmake if there’s no room on a server to host it after all, so the server configuration might have its own queue at times as well.)