SBMM Should Consider Hours Played & Perks Unlocked
Considering idlers and automated bots are a minority of the player base, then the majority can be generally categorized into <200h, <500h, <1000h and above 1k hours. Then the algorithm can mix and match the players from these categories, giving priority to players within the same play time range instead of purely focusing on "skill" as a metric. This should be coupled with % of characters/perks the player has unlocked, because after 1k hours I don't want to play with players that are ranking up their first survivor and have minimum knowledge of killer/survivor perks.
Similarly if the killer is "rank 1", then higher preference should be given to 1k+ hours survivors.
What I believe is the current case is that Behaviour only started collecting metrics and data recently, and the data pipelines are still not fine tuned for proper match making. In essence it must be an ongoing experimentation in the data science team.
Moreover, given the rise of new hacks that make survivors fly around and do gens instantly, it seems that behaviour isn't doing much on the server side to begin with, let alone collect meaningful data to caliber the player "skill".
Comments
-
SBMM should be considering the high and low ratings in lobbies, along with the perks/add-ons/items used. It doesn't really matter what you've unlocked. I have Ruin Haunted Undying Nurse but I shouldn't get punished for that when I play Pig lol
0 -
There are too many factors to look at. Besides perks they would have to look at items and killer add-ons.
0
