View Single Post

phalczen's Avatar


phalczen
04.26.2019 , 02:52 PM | #4
Quote: Originally Posted by Shariva View Post
Not based on the current kept statistics or ratings. The current tracking of statistics and rating that we all see in the "Battle Record" window, is based on statistics earned in GSF as it is or was. Since my return, I have, in example, my win/loss ratio seen going down from 75% to 43%. How did this get so bad? Well, imagine going up against a group each and every time. I'm not the best pilot, but I'm not the worst either. But the game makes the groups based on those same statistics and then when enough time has gone by, will pitch a ragtag band of lonewolfers against an organised group of pilots.
That's not correct.
Teams are organized based on the following factors:
  1. Games played across your entire legacy
  2. the total amount of requisition currently invested in the five ships on your bar
  3. How long you have been in queue
  4. If you are in a group or not
Win/loss rating does not contribute to matchmaking currently, as far as we can tell and as far as Bret Hoffman has described.

It has been hypothesized, by Drakolich, that there may be a "cap" to the number of games played across the legacy, such that a player with 1000 games played is considered equivalent to someone with 10000 games played.

Bullet point 2 has a couple of implications: first, that players who have a filled 5-ship bar will be considered higher "skilled" players in terms of matchmaking than players with fewer than 5 ships; and that players who have completed ships, meaning they have upgraded every single component available on that ship to max, will also be considered "higher" in skill than those who have simply mastered a ship (upgraded at least one component to maximum).
Spoiler


Assuming no one is grouped, if people in queue have been waiting a while and equally long, but are a wide variety of "skill" levels, then the matchmaker will construct teams as best as it can based on bullets 1 and 2.

I believe that it constructs the teams based on the algorithm first, and then sends out the invites. This is what results in a match that ends early for lack of players, because it can't predict who is going to decline the pop.

Lastly, with 5.10.2 I think, the system was changed to prioritize 8v8 over 12v12. Previously it had always prioritized 12v12, but with cross faction matches in 5.9 the population increased such that we saw a lot more 12v12 than we ever had. I interpret that patch note to mean that if there are 17-23 people in queue, the game is more likely to go ahead and pop an 8v8 if it has decent distribution of "Skill" rather than wait till there are 24 people in queue.

Obviously, true skill is more than just a measure or function of bullet points 1 and 2, but that's what we know about matchmaker post 5.9
If you think I've made a good contribution with this post, I kindly ask that you use my Refer a Friend link! Here is more information about the program.