Eureka! I've solved it - How to fix ranked play!

I have read every post complaining about ranked play. I have listened to every pro player and every streamer. I have debated every pleb on the forums, and have finally arrived at The Solution to fix ranked play. There is still some minutiae that needs to be figured out but here is largely how it will be done. The Solution should not take the Devs that long to complete. Hopefully, we can get an update on progress from the Devs Monday morning. I ask nothing for my contribution to the AOE2 community just remembrance that this is the day I single- handedly saved AoE2:DE.

The Solution

  1. Increase map pool to somewhere around 16-20 maps. For the time being these are the core maps that should be included: Arabia, Black Forest, Arena, Nomad, Gold Rush, Steppe, Ghost Lake, Oasis, Team Acropolis, Chaos Pit, Islands, Green Arabia and others TBD

  2. Instead of banning maps, each player or party lead will select the maps they are willing to play. This will be called “positive selection.”

  3. Here is the key - every map you select will impact an ELO multiplier. Nothing else about how ELO is calculated will change. You will still lose the same points, but if you only selected one map you will only receive 1/(total amount of maps) * traditional ELO points awarded for win. For example, if I only select Arabia out of 20 maps and win the game where I would normally be awarded 20 points, I will only receive 1/20 * 20 or 1 point.

  4. Above the map selection in the MatchMaking screen add some text along the lines of - “Note: By selecting more maps, you will increase the total points you can potentially win in your match.”

Benefits:

  1. This is the greatest mix of player freedom and map variety of any other solution. From what I can tell the community is greatly split on the forced map pool. Many players like it while many players hate it. This solution should appease all parties.

  2. While AOE2 is the greatest RTS of all times, it can be extremely daunting for new players. Every map lends itself to different strategies which can take a long time to get decent at. By allowing new players to slowly expand their map selection, it allows them to get comfortable with new maps on their terms and to join in on the fun of ranked play! The current system really repels new players.

  3. Because of the ELO multiplier, all players will be highly incentivized to play on as many maps as they can stomach leading to a great variety of games.

  4. Niche players who only like one map or legends like Blue Coffee and FatSlob will be able to play in a competitive mode with their settings. They will likely never see their ratings rise too high, but at least they get competitive games on their terms. The fact that this allows for legends to even exist is great and add to the wonderful diversity for AoE2.

  5. There should be no difference in wait times for players who are selecting many maps. In fact, this will likely unite the huge portion of the community that is still playing and Voobly so, theoretically, matching should take shorter.

  6. We won’t have to deal with constant squabbling over the map pool and which maps should be included.

  7. This is the perfect marriage between freedom of choice, quick match making, and increased map variety of ranked games. There’s no downsides. The current system alienates a huge population of the AoE2 community and risks losing many players. No player will truly feel alienated by The Solution and this should largely unite the communities.

takes a bow

Faithfully Yours,
FasteningCat7

6 Likes

No detriments? yes you must be the chosen one.

2 Likes

It’s an interesting idea to both allow to select a few maps while incentivizing to make a broad selectoin.
However, there are a few issues I see here that should be discussed:

  • Even when a player selects most of the maps he still might have a longer waiting time because others at his Elo level did not select any of them.
  • Some players selecting only a few maps will not understand that this impacts their waiting time and might get frustrated and leave.
  • Players selecing only a small subset of the maps will take much longer to reach an Elo that is according to their skill level on the maps they play. As matchmaking is based on Elo, the games will not be balanced during this time, and it will be an unsatisfying experience for these players and their opponents.
  • How many points will be detracted for players who only select a few maps?
    In case only the same fraction of the Elo is detracted, these players will eventually reach (roughly) the same Elo as they would otherwise - just taking many more games.
    If they do not only lose a fraction of the points when defeated, they will be ranked lower in the ladder compared to their skill level on the map they play, because their losses have a disproportionate impact on their Elo. In the latter case, this again would result in unbalanced and unsatiesfying matches involving these players.
  • When there are few players at some skill level queued up (particularly at the very top or bottom of the ladder), they might (try to) force maps on their oponents. Assume that player A and B are the only players of their Elo-level queued up (e.g. both at the very top of the ladder). Assume that A thinks he has an advantage when playing on, say, Team Islands. Now, when A is selecting only this map, no game will start unless the B also selects the Team Islands map. In other words, now the “meta game” will include on whether you can force some particular map on your opponent, e.g. by having more patience when waiting for a game, until your future opponent finally gives in by also selecting the map you want to play. This might be alleviated to some extent by reducing the Elo gain that can be achieved when selecting only a few maps, though.
6 Likes

win 1v1 arabia game with one map selected = wins 1-2 points
lose the same 1v1 arabia game = lose 10 to 20 points
I fell like this math makes smurfing easier - and promotes lots of elo loss.
Imagine, generic pro playing only 1v1 arabia, losing once every 10 games, and still getting closer and closer to zero elo and bashing more and more noobs there.
thankfully, @WoolMacaroon163 managed to develop the idea better than i.

3 Likes

Responses below. When you bring up minor non-issues, keep in mind that the current system is totally unbearable for potentially half of the AoE2 community and The Solution is a huge improvement. See responses below.

Ultimately, The Solution is a compromise. It appeases a huge proportion of the community that can’t currently stand the map pool. Nothing will appease everyone, but this is the best and most fair option.

1 Like

First, can you even smurf on Steam? Aren’t you stuck with one username unless you buy another version of the game on a different Steam or Xbox account? Smurfing is when lower ranked players specifically target higher ranked players to try to beat them without risking their rating to have bragging rights. MatchMaking in general makes smurfing impossible.

Ignoring that, you’re suggesting that someone will stay at a low ELO so they can just beat people more easily and not care about their rank? What’s to stop someone from doing that today? Why can’t I play 1v1s and just resign after 1 min if I get a map other than Arabia. That way when I get Arabia I can crush noobs? Sounds like a really dumb idea.

There will be some map specialist who only play on one map and will have lower ELOs on that map, but that’s part of the compromise and they are incentivized to play on more maps. See my other response for more details. If this is the worst past about The Solution, it’s definitely way better than what’s currently in place. I’d much rather play better players on maps I like than to be forced to play maps I hate. Many players like myself will not even play DE because of the map pool.

Have I answered all your concerns and made you a believer?

2 Likes

Simple. After 20 seconds add a popup, where you can click “uhm… I don’t really care that much, just lemme play already” or “no, I’l be patient”.

Is it not the other way around? A pro getting a low ranked account to bash noobs is what i know to be smurfing - unless i’m really wrong there. I think my clarification here makes my argument clearer.

Good point. CS GO and some other games will ban you if you try doing that - i have friends who got bans of over a week for quiting games in CS and Dota.
What i meant is that the player will bash noobs without wanting to bash noobs - he loses too much elo on defeat, wins too few elo on victory, ends up locked in that point with a bunch of lower levels. This would totally happen to me, as i only played 3 ranked team games in my entire life, and all my games (including 1v1) were on arabia (or oasis, when i failed to read the lobby description)

I must say your ideas are good. My concern is the way elo is calculated - part of your math, but mainly the AoE math - it doesn’t even take incompetent teammates in consideration! Perhaps flagging the dude as a “1v1 arabia only”, or a “any map, any team, any gamemode” (similar to integrity levels a lot of competitive games have) can work better than direct elo hits. Doing that is awfully hard, tho (in a 20yo ranked game without support to anything, even more).

As for smurfing, sadly there is no way to stop that. A motivated troll always finds a way. Sometimes he has to buy a new account, and sometimes one hour of losing in purpose gets the job done. But let’s try to get tryhard trolls out of the picture, for the sake of argument.

2 Likes

The OP’s idea violates the ELO rating system. One basic rule of an ELO system is that upon finishing matches, ELO points are only distributed, no points are completely lost and no new points are added. This guarantees that the average ELO stays constant over the complete player base such that the ‘value’ of the ELO ‘currency’ stays constant, and ratings from one date can be compared to ratings of another date. In the supposed system, the average ELO declines towards zero over time, which means that matchmaking will fail to make fair teams more and more.

I’m trying to think of a variation to the proposed idea that satisfies the ELO system. Say that we use the normal system as a baseline that prescribes the ELO everyone should gain or lose, and change these amounts by the ratio of the number of picked maps, based on who won. I propose the following system for a 1v1 match:

  • E = the number of points tranferred from loser to winner in the standard MM system.
  • N = number of maps picked by player 1.
  • M = number of maps picked by player 2.
  • C = a fixed constant that determines the influence of the amount of maps picked on the ELO change.

Then the amount of ELO points E’ to transfer from the loser to the winner should be:

  • E’ = (N+C)/(M+C)*E, when player 1 wins
  • E’ = (M+C)/(N+C)*E, when player 2 wins

For example:

  • Suppose C=10 and a baseline of 10 ELO points transferred (E=10) and players pick 2 resp. 14 maps (N=2, M=14). If player 1 wins, then E’ = (2+10)/(14+10)*10 = 5, so five points are transferred from player 2 to player 1. If player 2 would have won, then it would have been the other way around, so (14+10)/(2+10)*10 = 20 points transferred. So because player 2 picked more maps, his potential win is larger than his potential loss, and for player 1 the other way around. If you think this difference is too large, then simply increase the value of C, then the potential win and loss amounts will be closer together.
6 Likes

While I tried to keep things simple, I’d be fine with some other mathematical variations to the ELO formula as long as it accomplishes the same thing. I do like your idea as it solves for the issue of map specialists who have lower ELO than their true skill. Losing to them would be less punitive from a points perspective. Let’s say there were people who only played Arena. If you didn’t unselect Arena at least you wouldn’t lose as many points if you got matched against one of them.

I think this could be in The Solution 2.0. I don’t want to over complicate it now as I’m just trying to garner broad support and to do that, it needs to be simple. Also, I don’t think we have to be purists when it comes to ELO. Naturally the top of the ladder will constantly drift higher. I’d be surprised if they don’t allow for people to reset their ratings (maybe by uninstalling and reinstalling) and there’s always new players so the ELO currency is always being cheapened. I’m more concerned with making the game fun for all than having a purist ELO system.

Thanks for this thoughtful idea!!

You have put a lot of thought into it, however the main issue is that you are viewing ELO rating as a ‘trophy’ instead of an internal tool that is needed to pair people with similar playing ability. “Penalizing” for picking only a few maps simply makes a higher skilled player match with lower skilled players, which is game breaking and messes up ranked play completely. Ideally they should hide a players rating altogether so the ‘trophy’ aspect is removed completely and people have no idea what their rating is, but I don’t think people would stand for that. (Egos are real). Some people will be better on some maps, some people will do better with certain civs. Do you want a different rating for each civ as well?

The best idea in my opinion is to 1) restore the old initial release map pool for ranked play, where the maps are not so different that it will skew ratings entirely, and is closest to ‘standard’ play (i.e. no one is going to ‘hate’ ghost lake but love arabia). 2) Then allow skill ratings to be viewed in non ranked play lobbies, where people have the freedom to decide all the variables they want and play less conventional map types. I’m not even sure team games should be in ranked at all, after all having a teammate massively impacts the outcome of the game.

2 Likes

Although the idea works, your reaction makes me realize that my proposed variant will still lead to skewed ELO’s. People who pick few maps will have an ELO lower than their skill level and players who pick all maps will have a too high ELO. I think the effect must be quite small (high value of C) before it holds that having a higher ELO correctly reflects the increased skill of being able to play on more map types. If the effect is small, it will not get out of hand too much, because it is counterbalanced by the standard ELO-rule that you can win less than you can lose when playing against a much lower rated player.

Resetting ratings violates the ELO system again because of points being added or removed. I see only downsides to resetting the ELO to the starting value, so no reason to introduce that. People wanting a fresh start can always ‘smurf’ by making a new account.

New players entering the game does not devaluate ELO points. The average stays at 1000 points per player, so an ELO of 1000 will be an average player, regardless of whether you have it now or that a new player has it when entering this game a few years from now.

Leave queue an rejoin queue with different settings.

I love increasing the map pool, just use a scroll bar, doesn’t matter if it looks unsophisticated.
But the problem this creates is specialists that are really good a 1 map will keep a low rating.
This is not fun to play against if you’re a generalist.
On that 1 map the specialist will be a lot better than you.

1 Like

This is a non-issue. Most players who specialize will be Arena of BF players. No one loves Arabia and hates Gold Rush. If you don’t want to play against Arena or BF specialists you can de select it. I would. Also, most Arena and BF specialists are noobs anyways so people with lower ELO should be able to beat them. The specialist will all be incentivized to play more maps to find games more quickly and for better ELO. This is not an issue. I’ve thought of everything.

Can I get a TLDR?

Please and thank you

1 Like

We are encouraging changing ranked play be adding more maps and allowing for positive selection of maps instead of banning. Through ELO, players will be incentivized to select as many maps as they could like. This will be beneficial for new players and a huge portion of the community who hate being forced to play maps they hate like Arena, BF, and Islands.

I wish it didn’t take forever to que.

1 Like

Current match-making does not use real ELO, only an adjusted version of it which is vastly different. The same goes for Voobly since it developed the original ELO further some years ago. At this stage points are currently lost and created from thin air for both games, you will not get your ideal situation as this is not chess.

1 Like

I kinda like the solution, eureka indeed!

1 Like