I really think ranked auto-match should be separated from teamers much more than it is already. I don't see why for example I should get a loss counted on my record for playing against a set team with level 1 and 2's as teammates its frankly not fair or even fun.. I know its counted separately for the teamer but there is nothing of the sort for the players who are playing ranked automatch properly.
Would be interested in hearing peoples opinions on this one.
Keeping Teamers Separated In Automatch From True Randoms
3 Dec 2019, 18:07 PM
#1
Posts: 232
3 Dec 2019, 18:55 PM
#2
Posts: 8154 | Subs: 2
Whenever this games achieves 15 to 20 times the current playerbase.
My controls weren't working
And if they were, you were playing dishonorably,
And if you weren't, you were playing without skill,
And if you were, it's not fun to play that way,
And if it is, you only care about winning.
Bible of Scrublord, Psalt 3:16.
And if they were, you were playing dishonorably,
And if you weren't, you were playing without skill,
And if you were, it's not fun to play that way,
And if it is, you only care about winning.
Bible of Scrublord, Psalt 3:16.
3 Dec 2019, 19:24 PM
#3
Posts: 3260
The ELO system already handles this: if being in an arranged team makes you win more, your ELO goes up and you get pitted against individually more competent solo queuers.
The real reason for unbalanced matches is the small playerbase: the matchmaker can't give you equally skilled opponents and teammates if there aren't equally skilled opponents and teammates searching at that time.
So it does the best with what it has.
The real reason for unbalanced matches is the small playerbase: the matchmaker can't give you equally skilled opponents and teammates if there aren't equally skilled opponents and teammates searching at that time.
So it does the best with what it has.
3 Dec 2019, 20:15 PM
#4
Posts: 2145 | Subs: 2
The problem is each NEW team has no rank. So you get great premade teams smashing low rank players.
The matchmaker needs to try to gather a sampling of ranks for each team player. But there is no easy way to do it.
It needs to look for any rank in that mode (or similar mode) and use that as a starting point.
So if playing 3v3, ranks in 3v3 should be averaged for each player then averaged among them to get a base ELO. If no 3v3 ranks are found then 2v2 and 4v4 could be averaged.
But many people only play on premade teams so this will not work for them.
At this point the server would need to search all similar mode team ranks and maybe take 50% of the team rank as a starting point.
The best suggestion I have heard is from Duffman, where the team suggests what rank they want to start getting games at. And the server adjusts from that starting point. Because getting 5 minute noob smash games is not fun for good players either.
The matchmaker needs to try to gather a sampling of ranks for each team player. But there is no easy way to do it.
It needs to look for any rank in that mode (or similar mode) and use that as a starting point.
So if playing 3v3, ranks in 3v3 should be averaged for each player then averaged among them to get a base ELO. If no 3v3 ranks are found then 2v2 and 4v4 could be averaged.
But many people only play on premade teams so this will not work for them.
At this point the server would need to search all similar mode team ranks and maybe take 50% of the team rank as a starting point.
The best suggestion I have heard is from Duffman, where the team suggests what rank they want to start getting games at. And the server adjusts from that starting point. Because getting 5 minute noob smash games is not fun for good players either.
3 Dec 2019, 20:18 PM
#5
Posts: 232
Personally I think the best solution to this is to remove teams from ranked auto-match altogether and just have them in custom games instead.
OR have a team automatch..
OR have a team automatch..
3 Dec 2019, 22:14 PM
#6
Posts: 960
The playerbase is simply too small for that kind of change, and forcing pre-mades out of automatch entirely would be completely unfair to those players. Even SC2 at its peak mixed pre-mades and randoms.
The only solution I can think of is making the match-maker MUCH more strict when matching a pre-made with randoms, so you don't end up with massive discrepancies in skill that destroy any hope for the random team. Ideally, when searching as a group, your opponent shouldn't be more than 100-200 positions lower than you on the ladder; unfortunately, I get a lot of games where the difference is 500-2000.
It also happens at the top end; even the best 'random' team is going to get destroyed by a decent pre-made. Good coordination is extremely powerful, and can make up for a surprisingly large skill-gap.
The only solution I can think of is making the match-maker MUCH more strict when matching a pre-made with randoms, so you don't end up with massive discrepancies in skill that destroy any hope for the random team. Ideally, when searching as a group, your opponent shouldn't be more than 100-200 positions lower than you on the ladder; unfortunately, I get a lot of games where the difference is 500-2000.
The ELO system already handles this: if being in an arranged team makes you win more, your ELO goes up and you get pitted against individually more competent solo queuers
It also happens at the top end; even the best 'random' team is going to get destroyed by a decent pre-made. Good coordination is extremely powerful, and can make up for a surprisingly large skill-gap.
Opinions reflect all team modes (not 1v1)
I often edit my posts several times after posting
I often edit my posts several times after posting
3 Dec 2019, 22:55 PM
#7
Posts: 232
The playerbase is simply too small for that kind of change, and forcing pre-mades out of automatch entirely would be completely unfair to those players. Even SC2 at its peak mixed pre-mades and randoms.
The only solution I can think of is making the match-maker MUCH more strict when matching a pre-made with randoms, so you don't end up with massive discrepancies in skill that destroy any hope for the random team. Ideally, when searching as a group, your opponent shouldn't be more than 100-200 positions lower than you on the ladder; unfortunately, I get a lot of games where the difference is 500-2000.
It also happens at the top end; even the best 'random' team is going to get destroyed by a decent pre-made. Good coordination is extremely powerful, and can make up for a surprisingly large skill-gap.
I see where you are coming from , but it's not unfair for everyone in a ranked matched to play the roulette wheel on an equal footing rather than some people already knowing what numbers they will get before the spin.. It's actually a lot MORE fair for everyone involved to have an equal footing and then let the ELO do its thing.
3 Dec 2019, 23:10 PM
#8
Posts: 3032 | Subs: 3
The only problem is that you get massive ELO punishments if you lose vs an unranked team (less than 10 matches played together), even if this team consists of top players only.
Other than that: Leave it as it is, or else you're gonna have 30 minute queue times as arranged team. It is also a very big satisfaction if you win vs a decent arranged team with random teammates, and I dont want that to be removed from the game either
Other than that: Leave it as it is, or else you're gonna have 30 minute queue times as arranged team. It is also a very big satisfaction if you win vs a decent arranged team with random teammates, and I dont want that to be removed from the game either
3 Dec 2019, 23:16 PM
#9
Posts: 960
I see where you are coming from , but it's not unfair for everyone in a ranked matched to play the roulette wheel on an equal footing rather than some people already knowing what numbers they will get before the spin.. It's actually a lot MORE fair for everyone involved to have an equal footing and then let the ELO do its thing.
Neither situation is ideal. I've had days/nights where the larger modes have become literally unplayable, because it kept matching me with the exact same high-skill 4player group, which I had no chance of beating with random players. But on the other hand, that high skill group paid for the same game I did, so they shouldn't be forced to wait ages just to play with their friends.
I've also been in the opposite position, where I've had a 4player group of new players; getting them to stick with it if the queues were 30+ min would have been impossible.
The only problem is that you get massive ELO punishments if you lose vs an unranked team (less than 10 matches played together), even if this team consists of top players only.
I think this is the much bigger problem. Pre-made teams should probably use the average of the player's skill with their currently selection factions when match-making, rather than the current system which essentially assumes none of them have ever played before.
Opinions reflect all team modes (not 1v1)
I often edit my posts several times after posting
I often edit my posts several times after posting
4 Dec 2019, 01:59 AM
#10
Posts: 3423 | Subs: 1
I mean that's the risk you take with random teammates. This has nothing to do with arranged teams and everything to do with the amount of people who play
It can be just as frustrating to get paired with a bad teammate while playing against other randoms. You're not going to stop getting bad teammates just because we remove arranged teams
It can be just as frustrating to get paired with a bad teammate while playing against other randoms. You're not going to stop getting bad teammates just because we remove arranged teams
4 Dec 2019, 22:51 PM
#11
Posts: 232
I mean that's the risk you take with random teammates. This has nothing to do with arranged teams and everything to do with the amount of people who play
It can be just as frustrating to get paired with a bad teammate while playing against other randoms. You're not going to stop getting bad teammates just because we remove arranged teams
You are utterly missing the point. The other side will not be able to choose their teammate and must play the ELO roulette like everyone else ..
4 Dec 2019, 23:04 PM
#12
Posts: 607
How would a separate system even work for say a team of 2 wanting to play a 4v4? A separate ladder there or just non-ranked play?
TBH when I read the thread title I assumed it was about how you aren't guaranteed to spawn next to your arranged team mate in such situations.
TBH when I read the thread title I assumed it was about how you aren't guaranteed to spawn next to your arranged team mate in such situations.
4 Dec 2019, 23:08 PM
#13
Posts: 3588 | Subs: 3
And ranks in teamgames of 4v4 and 3v3 matter because: .....
4 Dec 2019, 23:08 PM
#14
Posts: 203
Having friends OP, plz nerf friends.
You can't spell "Power Creep" without REEEEE.
4 Dec 2019, 23:09 PM
#15
Posts: 17914 | Subs: 8
And ranks in teamgames of 4v4 and 3v3 matter because: .....
If you play 4v4 in a team with just 1 player, you get 2v2 ranking.
Maybe it's 'cause the Axis players aren't a bunch of forum crybabies, and they keep their cool and fight the uphill battle, unlike some of their "allied" counterparts.
4 Dec 2019, 23:09 PM
#16
Posts: 3588 | Subs: 3
If you play 4v4 in a team with just 1 player, you get 2v2 ranking.
Fair, but generally I would avoid that because the 4v4 playerbase is uhhhh mentally unstable.
4 Dec 2019, 23:17 PM
#17
Posts: 607
Fair, but generally I would avoid that because the 4v4 playerbase is uhhhh mentally unstable.
That's your choice, but plenty of people like to play big silly games with lots of explosions because it's fun; even with the badly typed insult fiesta that is 4v4 randoms. There's a different experience and is totally different from what 2v2 or 1v1 can ever approach in terms of spectacle, mayhem, and sheer scope.
My friend and I were rank 30 in arranged team of 2 because we played 4v4s.
Our choice to have fun in big dumb games then led to us having some weird 2v2s; but also it really messed up the ranking in general.
Anyway, I brought up the question because it seemed most people like the idea of separating arranged from random in concept, and so I was curious how this hybrid situation would pan out.
4 Dec 2019, 23:50 PM
#18
Posts: 3588 | Subs: 3
That's your choice, but plenty of people like to play big silly games with lots of explosions because it's fun; even with the badly typed insult fiesta that is 4v4 randoms. There's a different experience and is totally different from what 2v2 or 1v1 can ever approach in terms of spectacle, mayhem, and sheer scope.
My friend and I were rank 30 in arranged team of 2 because we played 4v4s.
Our choice to have fun in big dumb games then led to us having some weird 2v2s; but also it really messed up the ranking in general.
Anyway, I brought up the question because it seemed most people like the idea of separating arranged from random in concept, and so I was curious how this hybrid situation would pan out.
By all means, I play 4v4 too, but I don’t understand why 4v4 random automatch’s ranks matter. Case of point my okw rank in 4v4s is around 4000 and my 1v1 is 134, so that speaks a lot on how unbalanced the game mode is.
I don’t think it’s a teamgame symptom though, because I find 3v3s a lot more balanced and fair, and in my view 3v3 rank is more indicative of player skill.
I disagree on premades being the culprit, I think it’s primarily smaller map sizes over player count, as well as being more prone to bugsplats while loading.
5 Dec 2019, 00:49 AM
#19
Posts: 607
By all means, I play 4v4 too, but I don’t understand why 4v4 random automatch’s ranks matter. Case of point my okw rank in 4v4s is around 4000 and my 1v1 is 134, so that speaks a lot on how unbalanced the game mode is.
I don’t think it’s a teamgame symptom though, because I find 3v3s a lot more balanced and fair, and in my view 3v3 rank is more indicative of player skill.
I disagree on premades being the culprit, I think it’s primarily smaller map sizes over player count, as well as being more prone to bugsplats while loading.
I'm not sure if I'm misunderstanding you or if you misunderstood me, but the ranking issue I brought up was that arranged team of 2 (2v2) is affected by arranged team of 2 (4v4) and how the two should not be conflated, which currently they are.
The debate about whether 4v4 random rank matters isn't something I care to pursue.
5 Dec 2019, 02:06 AM
#20
Posts: 3423 | Subs: 1
You are utterly missing the point. The other side will not be able to choose their teammate and must play the ELO roulette like everyone else ..
I understand how your idea works, there's nothing wrong with it concept. What I'm saying is in practice, it won't work, there's not enough people and there will still be other problems
1 user is browsing this thread:
1 guest
Livestreams
 |
|
|
26 | ||
 |
|
|
18 | ||
 |
|
|
112 | ||
|
|
|
|
6 | ||
 |
|
|
4 | ||
|
|
|
|
2 |
Ladders Top 10
-
#Steam AliasWL%Streak
- 1.831222.789+37
- 2.35057.860+15
- 3.1110614.644+11
- 4.921405.695+5
- 5.634229.735+8
- 6.276108.719+27
- 7.306114.729+2
- 8.262137.657+3
- 9.1045675.608+3
- 10.722440.621+4
Data provided by
Relic Entertainment
Relic Entertainment
Replay highlight
VS
-
cblanco ★ -
보드카 중대
-
VonManteuffel -
Heartless Jäger
Einhoven Country
Honor it
9
Download
1237
Board Info
900 users are online:
900 guests
1 post in the last 24h
6 posts in the last week
34 posts in the last month
6 posts in the last week
34 posts in the last month
Registered members: 49187
Welcome our newest member, manclubgayote
Most online: 2043 users on 29 Oct 2023, 01:04 AM
Welcome our newest member, manclubgayote
Most online: 2043 users on 29 Oct 2023, 01:04 AM