Yeah so if both ban each other’s map they have 0% and if they don’t then they have 100%, so in total they have something in-between 0% and 100%. Hence a probability… Except it’s not a coin flip here it’s the randomness of what the matchmaking will decide. Stars are aligned? You get matched against another player like you. Stars are not aligned? You get matched on Nomad, and now, with absolutely no way to prevent it.
You don’t do probabilities AFTER getting matched, you do them BEFORE. The player has no way of knowing against which type of player they’re getting matched. Another variable which renders the whole math useless is Alt-F4s when an unpopular map is chosen. Let’s take again the example of a guy who favorites BF. with the proposed system you think he has 50% of playing BF right? False, because in reality everyone is going to Alt-F4 / AFK when they see BF, players are not going to suddenly tolerate BF just because you decided to artificially remove their option to ban it, Viper literally forfeited a tournament game on BF you think the guys who play Arabia 24/7 will be any different?
I just had an even crazier idea, what if we had a system such that:
A player who favors Arabia will get the following ratio of games: 100% Arabia, 0% Arena, 0% Other
A player who favors Arena gets: 100% Arena, 0% Arabia, 0% Other
A player who favors Other gets: 100% Other, 0% Arabia, 0% Arena
What an utopia it would be right? Well guess what this is exactly what unlimited bans achieves, but I don’t see you advocating for this system for, hmm, reasons, probably because otherwise we would lose this incredible map variety that we see right now (Arabia 60% pickrate in 1v1), for sure that must be the fundamental problem with this solution, right? Surely if devs have decided to go for it, then there must be a reason right? For sure they wouldn’t force everyone to play in a lazy suboptimal system and timeout them when they try to bypass it even against AI, right?
I think having separate Arabia, Arena, and Other queues (with perhaps an opt-out for water maps in the Other queue) would be a pretty good improvement to the 1v1 ladder. Let people opt into as many of these queues as they want and perhaps also pick a favorite for them.
There are two maps that clearly are played the most in 1v1: Arabia then Arena. I don’t really see people wanting to play only Baltic 100% of the time, for example. So having a random map pack as a 3rd option seems like a good way to group all of the non-Arabia people together.
People make some interesting points here, but I think the map ban system is just fine. Maybe there could be a way to reward players who play more than the same map, though. So your additional ELO with wins would have a half life each time you win on the same map (i.e. 15 points, 7 points, 3.5 points, 2 points, 1 point for a player who keeps repeating the same map). That would still allow people to develop expertise on certain types of maps but also encourage them to branch out.
This will lead to all players playing Arabia and Arabia only, because it gives Arabia players the tools to avoid all other maps, whereas others have the option only in theory but not in practice due to the small size of the player base, especially at ELOs not too close to 1000.
I either don’t see that somebody wants Baltic 100% of the time. But I think it’s not too far-fetched to imagine someone has it as a favorite map and has the reasonable expectation to be able to play it every second game (in 1v1).
This entire discussion reminds me of… imagine you’re meeting once a week to play board games with friends. 5 of 8 people favor game A, two favor game B, one favors game C. Would you insist to play game A every single time? Or would you be willing to also play game C once every 8 weeks for the sake of the others? Same situation here, the only difference is that the ranked queue is a bit more anonymous and makes it easier to ignore that there are also people at the other end.
I guess the Arabia-only advocates solution would be to tell their friends: play either our game with us or look for new friends…
This is fair. I only used the AFTER probabilities to highlight that the BEFORE probabilities don’t tell the whole story. Like how the average number of children is 2.4 but no one has exactly 2.4 children.
It would be a utopia, but I don’t believe it’s possible with the current number of players or even perhaps from a technical perspective. They can’t even run a RM, DM and EW ladder at the same time.
I think the precedent makes this more tricky than your analogy.
If matchmaking with a map pool were in the game from the start, this would not be a discussion. It would be just “how the game has always been” and we’d be used to it. People who only wanted X settings would have gone to lobby.
However since the actual situation is that people used to be able to get whatever settings they wanted 100% of the time, and that’s changed* people are (rightfully) unhappy.
*the change is more than the Devs being mean though, it’s an entirely new system trying to get players games faster with standardised settings, players who otherwise would likely never have played against each other previously.
EDIT: For clarity with regards to your analogy it would be like this
A group of friends used to meet up in two groups and play separate games in different rooms. Now they can only use one room and has to play together. Everyone has to decide what to play, but neither group wants to play the others game.
I think your interpretation is that they should all come to a compromise
Another is that we should just look for two rooms again.
Completely agree! Although it sounds strange at first no map bans actually gives a lot of players more control over the maps they play. Atm oftentimes people end up playing nobody likes to play. For instance one player bans ara and the other arena while the map prefs are the other way around. So imo conclusion is either change mm to find matches based on maps (basically the infinite ban scenario) or remove map bans to begin with. Both approaches would be way better than how it currently is.
Fair point. I don’t have the voobly experience, but came back to the game after almost 20 years just because of seeing the ranked queue in some youtube video, and it seemed so much of an improvement over the very old days of the MS gaming zone. To me it seems very natural to have, under such a matchmaking system, the expectation of getting to play my preference at least 50% of the times in 1v1, 25% in 2v2, etc.
I do understand the viewpoint of players who would prefer a ranked lobby, the thing is just: once you accept a matchmaking system that gives you the tremendous advantage of not having to find opponents on your own, you need to give up a certain amount of freedom concerning what exactly you’re going to play, unless you decide that your preference should be the standard for everyone.
But that’s not even the point of the thread. That you can’t have your wish 100% if you don’t game the current system via ALT-F4 is clear. And it is also obvious that in an ideal world with an infinite player base we would have separate queues with separate ELOs for all combination of map and game modes, which would then instantly pair you with an equally skilled player that equally enjoys the same settings and all are happy.
The question that I raised is simply what should be done, in the imperfect world we live in, in the case of conflicting preferences.
Edit: One additional advantage of no map bans in that all these petty discussions about the proportions of open/closed/hybrid/water maps in the pool would immediately cease. Because without bans, the only thing that matters is that you find at least one map within the pool that you like and you get a fair playrate on that map.
The problem with the analogy is that it doesn’t reflect the way player preferences exist in Age of Empires. Game modes and maps are quite polarizing. A large segment of people who want to play BF 4v4s do not want to play Arabia at all. And the opposite is true too: many people who want to play Arabia don’t want to play BF.
A better analogy would be if you have three very different activities, say playing board games, going for long distance runs, and reading Shakespeare. We could try to play a board game one night, run 10 miles the next day, and read Hamlet on the weekend… but that will be horrible for everyone involved. The game’s matchmaking system needs to reflect the actual ways people play the game. If I like playing Baltic, and the matchmaking system puts me in a game with someone who like Arena, then it’s going to be a bad experience for both of us. It’s not fun to play with someone if they’re not having fun too.
You can look at Steam Charts and see that DE currently has more active players than there were at any point for the HD Edition (other than the start of covid restrictions, where both games jumped in popularity).
I find it rather ridiculous that having more players should result in less ability for players to find what they want to play.
That’s the entire point of a matchmaking system, isn’t it? To find friends who want to play the same things as you do? One of the key responsibility of the matchmaker is to filter out people who don’t want to play together. It already has a limited ability to do that: you can queue separately for 1v1s instead of TGs. You could make the same argument that people should compromise, only play 1v1s some of the time and be forced to play TGs because otherwise there wouldn’t be enough players. But that also would result in a horrible user experience.
Of course, of course. But it boils down to the question what is the “same things” are.
As you say, currently “same thing” is an aoe2 game with pre-defined game mode (EW or RM with conquest victory conditions, standard speed, 200pop and some other default settings I forgot) and fixed number of players (1v1 or team).
If you consider “same thing” to be even more narrow (including map and perhaps even civ choice (I heard calls for forced random civ in ranked matches already), matchmaking becomes impossible except for the overall most popular variant.
We should not forget in the discussion: the overall player number does not matter, what counts is the number of available players in the queue at a given ELO range and a given time. Because I believe (and hopefully you agree) that, in order to have fun, having opponents of approximately the same skill level is even more important than which map is played.
And that is what the system is currently doing: First players are matched according availability and ELO, and only afterwards the map is selected based on their preferences. The entire thread is only about the question: What happens if their preferences do not match?
In many ways I do disagree with that. I think people really underestimate just how polarizing different opinions of different maps are.
I think there’s only one behavior that makes sense in this situation: don’t match them together. I’m sure many players would rather wait longer to get a fun game than waste time playing a game that isn’t fun for them. That’s why people Alt+F4 when they see the map (or now delete all of their starting units): it’s because they would rather restart the game and spend more time waiting than play a match that the system should have filtered out.
But this (the match) is done anyway, also in the current system. My suggestion does not change that, but merely switches from playing a “compromise” map to picking one of the preferred maps.
In the end, my suggestion will lead to:
Arabia players getting to play Arabia more often
Arena players getting to play Arena more often
Nomad players getting to play Nomad more often
Water players getting to play Water maps more often
and the expense of
all players playing “compromise maps” (such as Gold rush) significantly less often.
very rarely players get maps they explicitly dislike
Of course, if your dislike for some map is much stronger than your preference for your favorite map, then the suggestion is bad for you.
I was hoping that for most people the enthusiasm for their favorite map is stronger than their dislike for their least favorite map. Perhaps this is a too optimistic assessment of the situation.
Yes, and that’s a huge problem with the current system. If, say, one player bans Arabia and favorites Arena, then another favorites Arena and bans Arabia, they still get matched together. Only it’s on whichever map they haven’t banned, so you end up in situations where the players get a match on Migration or some map that neither of them wants.
This is the point that I think gets lost in many of these discussions. Getting “matched” on a map you don’t want to play is often an absolutely horrible experience. That’s the main reason why Alt+F4 exists. We even see this in Aoe4 as well, where players have been Alt+F4ing when they see they get a water map (and from experience, playing hour long French mirrors on island maps where the wood runs out is really boring).
On the contrary, I think this is something that would happen extremely often in your proposal. You can see the play rates for various maps under different conditions: https://www.ageofempires.com/stats/ageiide-global
55% of 1v1s are on Arabia. there are clearly many players who want to play that map and explicitly dislike every other map. That’s why I suggested the “Arabia, Arena, Other” queues: there are strong preferences to play only Arabia or only Arena, but for the most part there aren’t dedicated 1v1 communities on other maps. From my experiences there are many players who ban Arabia simply to play a different map, and these people would use the “Other” category without queuing for Arabia. And I think this would be a better experience for them, since they wouldn’t have to deal with an Arabia-preferring player Alt+F4ing or not having fun on the map.
But this is exactly what my proposal avoids and improves over the current system. Instead of an Arabia+Arena pair has to play Migration, they get their favorite map with 50% probability. In the long run this evens out.
To be honest, there are many situations in life where the term “horrible experience” is more appropriate than in this one.
These players also complain about the current system. What I mean with “dislike” is not that you have one favorite map and do not want to play anything else, but rather having an aversion against very particular map for some reason. For instance Socotra (you don’t like the laming) or Black Forest (don’t like eternal games) or even Islands (H2O allergy). You can avoid these in the current ban system, but not under my proposal. This is the only true downside of my proposal. Again, not compared to a non-existing utopia matchmaking, but compared to the current way.
And that is exactly the reason why players who appreciate other maps rely on a working ranked queue to find some matches at all. The current one is acceptable, but it could be better. Hence the thread.
To avoid misunderstandings: Only because the preference of a player differs from Arabia and Arena it does not automatically mean that such a player is happy to play any map as long as it’s not Arabia or Arena. It should be pretty obvious that, e.g., Black forest and Socotra are attracting very different players who also have strong opinions about the respective other map.
And when someone doesn’t like laming on Socotra and doesn’t like eternal games on BF and doesn’t like islands and doesn’t like Nomad starts… that quickly adds up to the point where they really need to be able to pick Arabia or Arena if they want to.
No it doesn’t, but putting all of the people together into one pool would make it easier for them to find other like-minded players who don’t want Arabia or Arena. You can add a pick/ban system there, let people avoid water maps, etc. But the point is that Arabia and Arena are so dominant that it’s reasonable to group most other people into the “Other” category. And if people who want to play Arabia aren’t placed on water or other maps, then that removes a large source of Alt+F4s, so the people in that queue will have a better experience.
Less than a third of ranked 1v1s are played on maps other than Arabia or Arena.
I think any reasonable system needs to treat those two maps specially, since they distinctly are more popular than anything else.
Quickplay 100% arabia with quickplay arabia elo single and team
Quickplay 100% arena with quickplay arena elo single and team
Quickplay 100% BF with quickplay BF elo single and team
Ranked as proposed in this thread?
The reason I throw in BF is because if you look at play rates a huge chunk of unranked is BF games, and they deserve good balanced games with a matchmaking system.
If you treat these maps specially (separate queue), the play rates will gradually get more uneven until only Arabia (and with a bit luck Arena) are left. Why this is going to happen has been discussed many times in this very forum before, I will not repeat it again. If the general goal is to keep diversity, minorities must not be sidelined even further just because of being a minority. If they are, they’ll eventually vanish. You know it. I know it. The devs know it. This is why separate queues will never be implemented, as long as the general map diversity goal stands.
The objective of this thread was never to improve the experience of a minority who put their own preference over that of everyone else without any hope of compromise (i.e. the ALT-F4 faction). Why should I care about people who themselves don’t care about others?
My suggestions merely try to improve the experience of the majority of players who may have preferences, perhaps even strong ones, but are willing to make compromises once in a while, both for their own sake and the community as a whole.
At least in 1v1 bf is one of the least balanced maps out there. It’s popularity in unranked is die to tg and lower level players.
Well question is do we even have map diversity to begin with? Looking at play rates of maps I wouldn’t be too sure about that. And while I agree with your proposal that removing map bans would be a major improvement over current system I’m not sure this changes map diversity. Also we quite frequently have non competive maps in the pool. And diversity for the sake of it’s own can’t be the goal here (at least imo).
Why? Diversity does not mean that every needs to be played equally often, just that there are more options than just one or two.
What bothers me about the statistics above is that it’s a bit skewed. Maps other than Arabia and Arena are not in the map pool all the time. For instance, given statistics above, I assume Hideout is less played than Arena, but maybe just by a factor of two and not a factor of four.
It’s not supposed to do that. The diversity we have (number of maps that are offered) is fine in principle, although you always could argue about tiny details. What the suggestion is supposed to achieve is to replace matches on “compromise” maps with matches on actually favored maps, and, perhaps even more important, free the community from the decision-making process which maps to ban in order to increase the probability of getting their favorites.
I am not sure what competitive means in this context. I understand it only as a criteria for e-sport, which pertains only to a tiny fraction of the playerbase.
I think the goal is to give as many players as possible a decent chance of playing a map they like and to explore as many aspects of the game as possible.
We also shouldn’t forget that there is one building, an entire class of units, and several civ boni and unique techs that are completely irrelevant on Arabia and Arena. But they are part of the game, so it makes sense to ensure that they find some use.
