Login

russian armor

Random vs. Random, AT vs. AT

8 Apr 2014, 12:15 PM
#1
avatar of The_Riddler

Posts: 336

When I play 2v2 random I get matched versus top 25-50 2v2AT players about half the time (mates aren't always on, so random is a decent alternative). Next to that, it seems that you get matched versus the seperate random ranks of the AT players, thereby ignoring their AT rank, resulting in disproportionate ladder movements.


Problem 1: Randoms matched versus AT players.

(Problem 2: Randoms matched versus random rank of AT players, rather than AT rank.)


Solution: Seperate random players from AT players.
8 Apr 2014, 12:24 PM
#2
avatar of Katitof

Posts: 17914 | Subs: 8

Won't happen unless you want to see multiplayer games other then 1v1 dead or be matched against same guys whole evening.
Population is way too small.
8 Apr 2014, 12:33 PM
#3
avatar of The_Riddler

Posts: 336

The number of actively searching 3v3AT and 4v4AT teams are probably not very high indeed, but 2v2AT should not be a problem. CoH2 has the double amount of players when compared to CoH1, where it does not take very long to get a 2v2AT match in the latter.

Besides, a small population should never be an argument to mix random and AT, as players prefer longer waiting times over mismatches, in the long-run.
8 Apr 2014, 12:34 PM
#4
avatar of Cruzz

Posts: 1221 | Subs: 41

They could give an option for random players to disable getting matched to ATs. I honestly don't think forcing matches like these "because not enough players" has any positive effect in the long term, because these games aren't really that rewarding for the AT and are just frustrating as hell for the randoms.
8 Apr 2014, 13:17 PM
#5
avatar of Unshavenbackman

Posts: 680

There are a couple of threads on this subject already. http://www.coh2.org/topic/15079/random-2v2--why-reilc

Imo its a pain in the ass. I stopped playing just doing compstomps now and then. Iam hoping for more players after the western front. If that dont happen I think the game is dead.
8 Apr 2014, 13:20 PM
#6
avatar of akosi

Posts: 1734

Permanently Banned
There were matched At team against random half year ago, then why it wasnot a problem for you?

Anyway it is problem yeah,but that wont happen cuz low population..
8 Apr 2014, 13:26 PM
#7
avatar of Inverse
Coder Red Badge

Posts: 1679 | Subs: 5

There aren't enough players, it would take too long to find games. Relic's algorithm is likely smart enough to make sure the two random players have a higher average rating than the team in order to offset some of the advantage that comes with better communication.
8 Apr 2014, 13:36 PM
#8
avatar of Cyridius

Posts: 627

When I play 2v2 random I get matched versus top 25-50 2v2AT players about half the time (mates aren't always on, so random is a decent alternative). Next to that, it seems that you get matched versus the seperate random ranks of the AT players, thereby ignoring their AT rank, resulting in disproportionate ladder movements.


Problem 1: Randoms matched versus AT players.

(Problem 2: Randoms matched versus random rank of AT players, rather than AT rank.)


Solution: Seperate random players from AT players.


Enjoy waiting 30 minutes for a game.
8 Apr 2014, 13:37 PM
#9
avatar of The_Riddler

Posts: 336

jump backJump back to quoted post8 Apr 2014, 13:20 PMakosi
There were matched At team against random half year ago, then why it wasnot a problem for you?

Anyway it is problem yeah,but that wont happen cuz low population..


It was a problem back then as well, but AT ladders did not excist yet, so there was no way of knowing if you faced random players or AT players (they could have added eachother on steam after the game, for instance). Personally, I have not played CoH2 random in about 8 months, partially because I stopped playing overall. Therefore, I only just now report this.

As for 2v2, there are more than enough random players and 2v2AT players to seperate the two types. This mismatching problem is part of the reason people stop playing, whereby a downward spiral is created when the argument for not fixing it is a low population.

Not fixing obvious problems leads to a low population, not the other way around.
8 Apr 2014, 13:49 PM
#10
avatar of Cruzz

Posts: 1221 | Subs: 41

There aren't enough players, it would take too long to find games. Relic's algorithm is likely smart enough to make sure the two random players have a higher average rating than the team in order to offset some of the advantage that comes with better communication.


"Take too long to find games" is a really tired argument. While I obviously can't speak for everyone, I for one would rather wait 15 minutes for a good game than stomp a newbie after 3.

Also I wouldn't put much faith in Relic's algorithms when their handling of the new provisional ranking is obviously dumb as a brick (instant...and completely random matchups for 10 games in each game mode for everyone, and with each partner too in AT). For that matter, looking at what kinds of matches higher ranked ATs seem to get, I seriously doubt the matcher is doing anything except try to find players with a combined elo at most x percentage away from the AT, where x keeps growing until a match is made.
8 Apr 2014, 14:05 PM
#11
avatar of Inverse
Coder Red Badge

Posts: 1679 | Subs: 5

Well, your opinion is not indicative of the opinion of the community at large. Every developer of multiplayer games with matchmaking these days prioritizes timely matchups over better matchup balance. They wouldn't be doing this if they believed a majority of their player base was fine waiting two or three times as long if it meant a better matchup. People don't like waiting 5 or 10 minutes for a game. Look at what Valve has done recently with Dota 2 matchmaking. When they split off into ranked and unranked, they removed a number of options that segmented the player pool in order to keep wait times down around their 2 minute 30 second target. And this is a game with 400,000-600,000 concurrent players every single day. Even they don't want to risk segmenting their community too much.

As for the first 10 calibration games, those aren't completely random. You still have a rating, it just has a higher degree of uncertainty, so the range of players you can be matched against is greater and the number of points you can gain from a win is greater since the system doesn't have enough games to make an accurate guess about your skill. It's pretty standard procedure in matchmaking systems these days. So is weighting arranged teams higher than random teams.

As for higher ranked teams getting poor opponents, that's kind of how matchmaking works. It's awful for players at the top because there aren't enough opponents at their rating, so they wait longer and get weaker opponents more often than the average player. This is doubly true for a game like CoH2 with such a small base of players.
8 Apr 2014, 14:38 PM
#12
avatar of The_Riddler

Posts: 336

Well, your opinion is not indicative of the opinion of the community at large. Every developer of multiplayer games with matchmaking these days prioritizes timely matchups over better matchup balance. They wouldn't be doing this if they believed a majority of their player base was fine waiting two or three times as long if it meant a better matchup. People don't like waiting 5 or 10 minutes for a game. Look at what Valve has done recently with Dota 2 matchmaking. When they split off into ranked and unranked, they removed a number of options that segmented the player pool in order to keep wait times down around their 2 minute 30 second target. And this is a game with 400,000-600,000 concurrent players every single day. Even they don't want to risk segmenting their community too much.


If timely matchups are prioritized over better matchups, it does not mean there isn't any improvement in this tradeoff. Would a player be willing to wait an additional minute for a closer matchup? Perhaps. Would a player be willing to wait an additional 5 minutes for a perfect matchup? Perhaps not. Would a player abandon the game after relatively long waiting times? Probably not. Would a player abandon the game after multiple stomps? Probably.

Id rather compare CoH2 to CoH1 than to Dota2, for all the obvious reasons. This leads to the observation that reasonable fast and good matchmaking of 2v2s and 2v2ATs is possible with the current player base.

8 Apr 2014, 14:42 PM
#13
avatar of Inverse
Coder Red Badge

Posts: 1679 | Subs: 5

Except vCoH's 2v2 AT matchmaking was literally fucking horrendous. It was awful. Literally awful. You'd get maybe 4 teams searching at once if you were incredibly lucky, you'd have to wait 10 minutes for a game, and then you'd almost never get an even matchup. Not to mention 3v3/4v4 AT matchmaking, which was entirely dead.

The CoH2 approach to team matchmaking is infinitely better than that.
8 Apr 2014, 16:00 PM
#14
avatar of The_Riddler

Posts: 336

I was aiming at the current CoH1 (please stop saying vCoH, especially since nearly every 2v2 has OF factions) 2v2 and 2v2AT matchmaking. It isn't great, but manageable, therefore, a game with a much larger player base should be manageable as well.

Furthermore, all those random players should not be matched versus AT players, just because a few AT teams cannot bare to wait a few minutes. This is unacceptable. Moreover, the average skill cap between CoH2 players is smaller than between CoH1 players, mainly due to the life span of these games, making even AT matchups easier.
1 user is browsing this thread: 1 guest

Ladders Top 10

  • #
    Steam Alias
    W
    L
    %
    Streak
Data provided by Relic Relic Entertainment

Replay highlight

VS
  • U.S. Forces flag cblanco ★
  • The British Forces flag 보드카 중대
  • Oberkommando West flag VonManteuffel
  • Ostheer flag Heartless Jäger
uploaded by XXxxHeartlessxxXX

Board Info

969 users are online: 969 guests
0 post in the last 24h
8 posts in the last week
34 posts in the last month
Registered members: 49107
Welcome our newest member, Falac851
Most online: 2043 users on 29 Oct 2023, 01:04 AM