Rankings - let's talk about it

(Press E) #21

[quote=“watsyurdeal;c-225586”]1. If you guys are wanting to be rewarded for losing, but top fragging, I’m sorry but are we playing an objective based game or death match?

  1. Tell me more about how you balance matchmaking with a population as small as ours? Now do it for ranked, do you see the problem? It’a very likely the system is working fine but has too small of a player base to work with.

  2. Before you do anything…make sure you r gud, if not, git gud son. Be more like Amerika or Nail

And yes, I did skim through a lot[/quote]

You should consider going back and reading that then.

  1. Yes, it is an objective based game. But as I and others have said, as it is whether you win or not is mostly determined by getting lucky with a good team. I don’t think wins and losses should have no involvement in determining your rank, because I understand this game’s main goal. However, losing because you just happened to get paired with an incompetent team isn’t fun at all. Not to mention leavers, which this system would also lessen the impact of.

  2. Like I said, you use their own skill, not the luck they’ve had in being matched with teammates. MLG players get paired with MLG players, rather than MLG players being paired with players the system thinks are MLG because they’ve been boosted, or MLG players dragged down by terrible players. Of course the system would work better with a larger playerbase, but firstly, we don’t have one, and the devs need to base the game on what we have, not what we should have. Secondly, more players doesn’t guarantee balanced matches. When a system only ever takes team performance into account and teams can change at any time instead of staying fixed, these issues happen.

  3. Like I said, I’m not the best player ever. But I’ve been paired with a lot of teammates far worse than me, against teams far above ours. If my opinion isn’t valid because I’m not good enough, look to everyone else’s opinions here.

Basically my point is, good player gets paired with horrid team by pure chance, repeat a few times, get deranked a few times, and now the system is suddenly unbalanced, perceiving a good player as worse than they really are. This also works for terrible players being carried by good players.

(GatoCommodore) #22

sure playing with dimwit player in ranked is bad, but then this also happens in many games!
including:

DoTA 2
R6 Siege
Battlefield 1
Titanfall 2
Overwatch

note that these games are more populated than DB but still gets matchmaking problems

(at least from my experience, casual matchmaking in R6 Siege took about a minute or so, but still gets level 1s or 20s fighting enemies at level 170+)

matchmaking algorithm will only be fixed if bad players/newbies doesnt exist, which is impossible considering bad players will always exist

i dont know about the future tho when you can just download skill and muscle memory and tracking ability for your $1000 dollar prosthetic arms and eyes

until then, we just gotta bear with it

(StimmedGrunt) #23

Hey, new forum user, relatively older player, just at containment war. I do a bit of mathematics for fun, so if you gave me bounds and vars for what you would like to see in the new/improved system, give me 2 weeks and 15 bucks at Tim Horton’s, I could pump out a new algorithm for ya! I really only play PUBs myself, and have played 1 or 2 PUGs and sucked, but am willing to give it a shot! I’ll keep an eye out for postings, and my steam is CaffeinatedGrunt, with an orange haired dwarf sleeping on a shovel for a profile pic, so send anything my way!

WARNING, NERD ZONE NOW, ENTER AT OWN RISK!
As far as population is concerned, if you use a percentage based model with lots of repetition, you should be OK versus using ELO 400, which I believe is the closest if you truly can drop 4 ranks out of 22 in 4 games. CS:GO uses original ELO, with minor variance in personal(I think accuracy and K/D are factored) and DotA. I haven’t had experience with the Titanfall service, and don’t play too much siege, and don’t own B1 or OW. ELO 400 is a score that changes relative to 400 based on win performance over a course of 10, 100, and 1000 games(Rough translation). the only other comparison that I can see applicable here is the WN series of ratings, most popular in Wargamings simulators, World of Tanks and World of Warships.

(XavienX) #24

[quote=“watsyurdeal;c-225586”]1. If you guys are wanting to be rewarded for losing, but top fragging, I’m sorry but are we playing an objective based game or death match?

  1. Tell me more about how you balance matchmaking with a population as small as ours? Now do it for ranked, do you see the problem? It’a very likely the system is working fine but has too small of a player base to work with.

  2. Before you do anything…make sure you r gud, if not, git gud son. Be more like Amerika or Nail

And yes, I did skim through a lot[/quote]

Well most of the time as good as you are, you’re teammates simply are so inexperienced that even if you died once and got 100 kills, the other team does a good job holding up and are never all wiped out at the same time. That imposes you from advancing since you also don’t even have teammates helping you out besides killing themselves in spawn not knowing where the wasd keys are at.

(ThunderZsolt) #25

[quote=“watsyurdeal;c-225586”]1. If you guys are wanting to be rewarded for losing, but top fragging, I’m sorry but are we playing an objective based game or death match?
[/quote]

There is a huge difference between top fragging, and top scoring.
You can top score with a medic or engineer with 0 kills. And also looking at other stats like accuracy makes 100% possible to determine if you are better than your team or not - but the current system doesn’t care even if you have an AFK player.

Of course it is easier to say “git gud, the system is fine” but its clearly not when still about 20% of the ranked matches is finished 5v4 in silver ranking even after SD added leaving penalty (it was more than hal of the games before).