Something wrong with teamgames matchmaking.
you just can’t play casually with friends with a lower elo than you, becausein this case the system make sure you loses the game.
Instead of matching you with someone of your elo, and matching your friends with people of their elo, they match the average of your elo with the average with their elo.
So if you are 1300 elo and play with a 800 elo and a 900 elo, they match you with 3 players with 1000-1100. It results that none my teamates can’t beat anyone, while i end up 1vs3…
So i have a 25% winrates since i’m at 1300 elo playing with random people from a discord with this system… (while 70% winrates on 1v1 in platium).
It’s really annoying because it just forces you to refuse to play with beginners…
2 Likes
I think your complaint is justified. Sadly, this is also often how many games work. AoE is no exception in this. They average (team) ELO and see if they can match this average ELO versus a team with approximately the same ELO. This is for example as well how the Warcraft 3 Champions community ladder approaches this. They have some math guru over there and you are right what happens here:
A 1.300 + 700 ELO team vs a 1.000 + 1.000 ELO teams is favoured towards the latter. The math used has some specific names I cannot recall. But roughly speaking: ELO increase is not linear. Someone with 1.200 ELO isn’t twice as good as someone with 600 ELO. It is the absence of weighing in this variance in the match making system that causes these skews and biases in teams.
There are only two solutions: fix the math to account for this non linear distribution in ELO (which anyone with some math background can do, or even a statistician can approximate very accurately I guess). Or match people in teams with about equal ELO per player, not per team average. The latter will be more of a burden to queue times, but one could argue the extend of that.
This is interesting, in some games it’s the opposite and 1300+700 would slaughter 2x1000. I’m a bit surprised that doesn’t happen in AoE 4, I’d have thought the 1300 player could quickly take out one of the 1000 players then move on to the next one. In AoE 2 DE, I know a top player is able to easily beat multiple lower rated players, even when the total ELO of their opponents exceeds theirs.
In AoE 4 it is extremely hard to withstand two players who are of considerable lower skill. There’s just not enough micro to really make a difference and defensive buildings make it very hard to take someone out in the beginning which you usually rely on. You even need to build rams to finish him off which are quite costly and might be useless against the second player who should be in castle age by that time. And if you dont finish him off he just packs his entire villagers into his tc, towers, landmarks while his ally surpasses you economically. Team games with random people often ended up in me attacking the enemy wing, finally building rams trying to finish him off while in minute 15-20 his ally just reached castle age wiping out my units with supperior army.
I agree that current matchmaking results often in annoying matches… I often had the other example, where in 3v3: ca. 3x1200 against 2x1000+1600.
The 1600 ELO guy, could have played the 1v3 and still win…
I think, there is no easy solution for the issue… My initial thought was as well, to find a “peer” for everyone with similar ELO, but then in-game, there is no way to ensure that the 800 ELO player is not next to the 1600 ELO player… or maybe not next, but at so much skill difference, I am not sure what can help to make that match workable…
There’s an example here of Hera winning 1v3, I don’t know what rating the 3 players were:
Age of Empires 4 - Pro Gamer (Hera) 1vs3 - YouTube
it’s totally normal to win 2 lower skills.
depends on elo difference.
if you can not win, probably they is not “considerable lower skill”
they widen ELO range to reduce time to wait.
but TGs are BAD, people stopped playing and situation became worse.
just now played game, their opponents ram pushed on other flank.
If they would succeed → 5 players played 10 minutes for nothing.
Game became 3vs4.
Current system is not team games… it’s FFA with teams.
I don’t think that a 1300 ELO guy can win against two 1000 ELO guys. That includes that those 1000 ELO guys do have their strengths in teamplay meaning that they aim to not take a fight when their armies are not combined. And 300 ELO difference is a very big skill difference already to me.
Here’s Beastyqt playing against two 1000 (!) ELO players. Game did last 43 minutes which is a lot considering the skill gap was that large. Beasties ELO was probably >2000 at that time.
@breeminator
I’ve seen that video too. In my opinion it is not an appropriate example because those three guys did not really team up. Nobody helped the mongol player when he was under attack and hera played a clearly overpowered cheese strategy which got nerved a few days later.
Its made to ensure more fair environment for teams. What would happen if there wasn’t this system? You might end up facint 3x 1700 elo players while your friends might be under 1k.
If something the matchmaking is ■■■■■■ when they match ppl with too huge cap. Im around 1600-1700 elo but generally im forced to play with 700-1300 elo players so its complete coinflip for me what ever I can win or not. I play solely 2v2 and its not really fun to see when you got mate that has no idea what they’re doing and just dies to first 10 longbowmen that show at his base.
I want option to have strict matchmaking even if my queue times are 10mins because now they’re 5-8mins which has almost always been the case since I got to higher elos even when they made matchmaking changes and I can tell with 100% confidence that quality of my matches went from higher to trash tier.
Interesting game. I think the difference vs AoE 2 is the ease of executing the early build order. He comments on that at one point, and I’ve seen it before when top players match up against a significantly lower ELO player on the ladder. The weaker player can be fairly even for the early battles. In AoE 2, it’s much harder to execute a tight build order, where there’s always just barely enough food, and you need to push deer and lure boar.
I think those two players could probably have beaten him if they both went English, combined their early longbows, and focused on shutting down his eco from outside his TC range, rather than making rams and going for the TC.
Absolutely. I think not going for the TC could have made the difference here already.
He is simple explample how ■■■■■■ up things are.
Im the +1500 player and this is solo queue no premade for me.
Top of the cake my ally decided to play simcity instead of AOE4 which resulted him purely building base and nothing else. When I tried to push out and get map control I was facing 2ppl constantly and forced to turtle long time without getting anything done till I decided to leave
Here is the question if any developer or related person sees this. Who the ■■■■ thought this is fun experience?
Yea, I second that. Quickmatch team games with random team mates are horrible. Whenever I played quickmatch team games I did so with someone I knew. There’s just no point in risking to get mixed with someone who is of considerable lower skill level. And stomping two players who are of considerable lower skill level is a very hard task in this game.
Here we have got an ELO difference of 500 so it backs up my consumption very well.
you already were given example where Hera Win 1vs3
but you want to find an excuse why it’s “wrong example”
may be in the example above he tilted and resign on purpose? how can you made an conclusion without seeing the game.
Yes I really think that it is not an appropriate example as HA at that time was an overpowered unit, and hera went for mass HA mixed with a few other units. He would have probably also won had he gone with a mongol tower rush. Both strategies do not exist anymore as they were too strong. Out of the three, one player had a ranking of around 1 k ELO, I can’t find data for the remaining two players though (twitch users in that game also tried to identify them but couldn’t find something). So idk where exactly to rank them. Did hera mention their ELO in his game? Seems like they didn’t even play a total minimum of 10 1on1 quickmatches to get listed there.
Here’s another example of beastyqt playing against two ~1050 ELO players. He went for three relics reignitz cathedral, which just also got nerved a few days later.
With all those OP strategies being gone I would go the other way around: Do we have a current video of a pro player beating three 1k ELO players like hera supposedly did? Because if it was possible im quite sure youtubers like beastyqt would go for it.
