2.3k players are literally the fringe of the community. I mean, I’m pretty sure that if they wanted to play games without matchmaking they could notice each other’s presence online and work something out via talking to each other… oh wait, right, that important feature of the current lobby system is completely missing. This is the kind of issues they need to address.
Yes, sure, but you will only play games you aren’t actually having fun with so many times before you get sick of the whole thing altogether. If I had no option at all to dodge playing arabia 15 times a row, I would just not play this game.
Curious to know what the skill difference on Arena is between a 2k ELO Player, and a 2k ELO Player that only plays Arena. Which one will win?
From my perspective it seems they could be evenly matched. In which case, the ELO system would be working as intended - to help match players of similar skill levels.
From every instance of this sort of matchup happening in the past we know what happens.
Do you know who Modri is?
He knocked out BacT and ACCM enroute to the Semis, where he lost to Tatoh, who eventually won the tournament over Viper. He also went on to beat Hera in a showmatch after the same tournament, effectively claiming 3rd place.
He’s never qualified for a bracket stage in a non-Arena 1v1 tournament. That’s all you need to know about the difference between playing the same map forever and playing all the maps. The player who plays all the maps will be at a disadvantage to the player who’s done nothing but practice and develop within that map’s meta.
That’s why ELO is currently garbage. Hence, why making map picks the norm makes ratings absolutely, utterly, pointless. It’s already bad. Map picks would just make it a certainty.
we can clearly agree that the problem is mostly the current system of ranked people tey to bend so they are “forced” to alt f4 maps is secondary as there are alot of clones and it misses a bit variety but the main thing is that we should let the small Procentice of arabia only player or arena only people have their way in ruin the game expierence for people
If that’s a problem with the ratings system, then why not just display more information so players know how they perform on different maps? Keep an overall Elo, but display an Elo and win/loss numbers for the player on Arabia, Arena, or other common maps or “categories” of maps. That way you have a more detailed breakdown of player skill. Seems like a fairly simple band-aid that would help.
you can’t lose rating for dodging queue because elo has to be equalized (well… now that they finally fixed team game elo trading in the last patch). It wouldn’t go to anyone. Though maybe a rating penalty for dropping would actually be a way for them to drop those inflated elo points till the # in the system is correct again xp
moreso that the only time they’ll ever play anyone who isn’t an arena-only clown is when someone who plays all the maps plays him. Which puts that person at a distinct disadvantage. The more likely people are, and the more capable people are of playing one map, the greater the chances of that player will get a clearly, indisputably unfavored matchup.
You can, but unless such information is clearly present within the game, it’ll be totally useless. Further, that won’t somehow keep high win Arabia players away from high-win Arena players and the like. So long as Elo says 1740, it’s going to try to match you to 1740, whether or not your opponent has 700 arabia wins on DE, which is kinda the problem. We want ELO to provide us with balanced matches and as it currently is, it simply can’t do that unless we all decide ALT-F4 to one map and stay there. Only separating ELO by map type can do that. Which is what I’ve proposed.
If you separate ELO by map type, it actually doesn’t matter what map selection type is used. You can map pick and be matched up only against players at your skill level on the selected map, or you can go variety and play whatever and whichever map gets selected randomly will use that map type ELO to find you an opponent. Then ratings will be accurate, and therefore, matter.
Okay… so as I understand it the only thing wrong with an Arena player is that supposedly they have to cheat the system to play Arena?
Right. Okay. So what if they didn’t have to cheat the system and got more map bans?
As mentioned before, a person choosing to play all the maps has already been placing himself at a disadvantage. Therefore in theory he should be an equal match to the arena player if they were both 2k ELO
It could cause a “trap”, with players feeling like they have to specialize in one map or one strategy to win. Otherwise, if they go to a map they are not familiar with they just lose. So, separating ELO by map type could be a good solution. It would encourage players to try more maps outside their comfort zone, and they will be more likely to be matched against an equally skilled opponent. However, I’m not sure how complicated that will be to implement.
No. That was never a problem until matchmaking was created, because prior to DE, the only way to get a game with any setting is to host the game with the settings you want. Arena-only and Arabia-only players are carryover from that era, who thoroughly enjoy that concept.
The problem now is that in seeking a single map to play through matchmaking they are giving themselves an ELO that should be representative of their overall skill based upon the performance of their best map. This is the problem.
(aside from the one with players not getting to play the game because their opponent doesn’t like the map. That’s a problem too, but that’s short-term. ELO distortion is a long-term nightmare.)
The problem, which I’ve already explained is that if you give them more map bans, you will make the aforementioned ELO distortion worse and more common.
It’s even worse for a player who practices a first map, gets good, figures out the meta, climbs the ranks, then moves onto a new map (maybe a pool switch) and then gets a massive losing streak because they don’t know the meta as well. For that time period, neither the player nor his opponent is getting a quality game, or likely an enjoyable experience.
That’s why I mentioned Modri. In practice, the theory doesn’t work out. ELO simply can’t correct for that difference in maps. ELO hasn’t been designed to do so, it’s not even incorporated.
Just… make more ELO’s. If they can implement one, they can implement four. Open land, closed land, water, and hybrid. Roll Nomad into the “hybrid” category, then plug and play. It’s that simple.
Wouldnt that make it super boring and long to rank up though? Imagine in TGs. Its already taking players 100’s of games just to get to the average Elo. If you had to do that in 4 different map archtypes until you can find balanced matches, thats half a year of playing… Also would it encourage more players to simply stick to what they have already ranked up?
Or would the system translate some of the skill in one map to another and assume he should start his placements at the elo he has at his highest archtype to try and speed this up?
The system is already inaccurate so I guess even with even longer time to set up Elos in the long run it would work better per map. We should probs reset all Elos and do exponential Elo for first games to kick-start it If it were to happen. Elo right now is an indicator of nothing we need a reset anyway why not make the system more precise even with the awkward transition stage in long term would be better
Okay. Please elaborate as to how a 2k Arena player is any better than a 2k Random Map player.
You are right, the 2k Random Map player is at a disadvantage. However, he has already put himself at a disadvantage in order to reach 2k ELO in the first place. As I explained before…
Another point to make is that if people get more bans, you are right in that ELO will be distorted towards the settings the players pick. However… there needs to be a distinction between if ELO should be representative of a players true skill level, or if it should be used to help matchmaking decide on two similarly skilled opponents to face each other.
In my opinion, ELO should just be used by the system to match players against each other. So… who cares if the ELO is distorted for people that only want to play one map or one Civilization?
How about we rate elo per map but if you have multiple maps you play then you get the average Elo of all maps as your Elo used to matchmake with others? If you only play Arabia you get perfect matchmsking if you play all maps you still get a relatively close matchup
On arena. Strictly speaking, a 2k arena player is worse than a 2k random map player. That’s literally half my point you’re just casually ignoring for the sake of harping on this. That might be the most important part I’m trying to get through. What if they want to play something else for a bit? Then it’s a 2k Arena player vs a 2k pool of players who don’t have all their rating playing Arena on something that’s not arena and that’s a poor matchup to say the least. Vitally important caveat aside…
On arena? In practice. If you’ve seen any AOE 2 games, you know that certain players gain an utterly massive advantage on certain maps and are utterly clueless on others.
Or is the meme of Hera (the only 2600) just sending it on water maps not evidence enough that maps distort ELO balance to a hysterical level? If it’s not, I see absolutely nothing I can give you as an example or an explanation that would stick. Favoritism towards a map makes an absolutely massive difference in how the game plays out.
You’ve assumed that, and that’s not necessarily true. It’s very possible he’s had most of his matches against variety players, and the few games he’s had against Arena clowns were games where he outrated his opponent enough to overcome a meta disadvantage. A little under 15% of games played in 1v1 matchmaking are on Arena, and an even smaller proportion of those are Arena clowns, so a skilled player can still climb the ranks, even in spite of having a clear disadvantage on those couple games.
These two things you are setting against each other are one and the same. If ELO doesn’t properly evaluate a player’s true skill level, it’s totally useless for providing games between two equally skilled players. There are no other metrics that we use to make such a distinction.
If ELO is inaccurate there is no way to get an evenly matched game. That’s the point of fixing it.
Not sure why this thread isnt merged into The Alt+F4 Multiplayer Problem. To me it looks like both threads are about exatly the same issue. Dodging games is just equivalent to Alt F4. Can someone delight me and tell me the difference between these threads?