This raises a big issue for new players. You have to lose 10-20 games to get to your skill level for new players. Because elo 1000 players already know the game. Most people are not willing to lose 10-20 games before discovering their place so they quit.
The new player experience is bad in this game. Starting at 1000 elo is way too high for someone to start playing.
The starting elo should be made 500. It’s not fair. If people start at 1000, they’re new players, and then they’ll drop elo, when every game is a loss for the next 10-20 games you aren’t going to feel like playing. 500 elo is a much better new player elo starting point.
And when you actually hit 1000, it’s actually a milestone. It’s not like “oh, I lost 500 points and I just recover right now”, if feels like you make progress rather than you recover what you already lost.
When you start the game at elo 1000, you’re not a real 1000 elo player. You still can go down from 500, but you have a lot of places to go up.
It’s like starting at Gold 4 in League of Legends, makes for a terrible new player experience.
I started Bronze 1 in League of Legends after winning 7/10 placement matches. Currently I’m Platinum 2. And it’s great that I started at Bronze 1 because had I started at Gold IV I would have quit the game a long time ago.
The problem is how the elo system is set up to (very gradually) shift the average elo to be the starting elo. So eventually 1000 elo now would be equivalent to 500 elo. It’ll take a long time, but it will happen. With changes to how the elo system is calculated though, to prevent this, I could see it working.
Whatever the starting Elo you set the ladder it will be the new average Elo for all the players
It’s only matter of time
Set divisions on the ladder is a good idea if you have the player base of LoL
IMO the first 10 games should have a higher coefficient to put you in your Elo faster but also the game should have a better introduction to MP for a new players
It is a hard game to learn, special for new players that only know how to pull a trigger
Yeah, ngl part of the reason I don’t do ranked is the fact that I don’t feel like getting punched in the face 20+ times in a row first, so I just don’t even bother, even though the one ranked game I played was kinda fun.
Everybody lose games. There are many people have lost more than a thousand games already. The players should just quit if they can’t take 10 to 20 loss.
New players will not lose so many games if they got some basic knowledge such as unit counter and just 1 build order. All these can be done by watching video and play against AI.
BTW, it will be interesting if they set the starting elo to 0.
I mean there’s also a difference between 10-20 losses over a month long period with wins interspersed throughout too, and 10-20 losses in a row.
To an extent I agree, the person I played in my one game didn’t believe it was my first ranked game because of some of the decisions I made, which were things I learned from watching videos, but the number thrown around across the board is you’ll lose your first 10 regardless. That number as a minimum baseline of losses when starting ranked i’ve heard pretty consistently.
They should realize there is something wrong when they are in a loss streak. It might be not a good moment for games.
Only the people who have fun playing the game will keep playing. Loss is also a part of playing the game. I would say the game isn’t for them if they can’t have fun or learn something from loss.
Why not advocate for a better explanation for first timers on how ELO works?
Why waste time with void actions when one can do something that has impact indeed?
Why not copy-paste the formula from chess? Chess also has ELO and they don’t have this problem. ELO was introduced in 1960. No.1 in the world is currently 2900 (almost 3k) and the starting elo didn’t become the average elo in 90 years.
I think what you are suggesting is a void action. What point is an explanation when the system still sucks? I didn’t need an explanation in chess or league of legends and the ranking seemed fair. Reducing the starting elo from 1000 to 500 will have impact indeed.
firstly we need data to know what the average elo is that newcomers reach after their placement games (if it is 1000 or close to it then none of this is necessary)
if newcomers usually fall far below 1000, then i think it would be better to make a completely new system. maybe some kind of league where the top 10% go up to the next league at the end of the season or if they have a winrate above x% (and relegation of weakest/inactive players)
First of all: do you know how much ELO newcomers lose after a defeat during their first 10 matches?
Second: you can see what happen to ELO when you dont have a HUGE playerbase like chess and LOL. EW ELO doesnt grow too much, even to top player and DM TG ELO is getting lower to everybody. Decrease RM ELO to 500 and this will happen there too
I think the ELO system is fine. The problem is the lack of player base. If no one of your elo is online then you match with different elo.
Personnally I prefer to match with a higher elo than waiting a game for 30 min. And if I match with a lower elo, it’s not very challenging for me but the opponent will learn something )
First of all: Most new players I know (new players for real, not coming from AoE2 HD or older) lose 10/10 new player games. It’s not even close.
And anecdotal aside, it’s simple logic, given the AoE2’s ELO distribution, with 1000 ELO = average player; you don’t need a study to figure out that new players are solid below average.
Second, you confuse causality with correlation. ELO doesnt grow too much, because not many people play it, because the ELO system is already bad.
Most people playing AoE2 play either singleplayer or online casual. Why would they stay away from rank? Because they don’t like losing game after game after game.
ELO system itself is fine, it’s been used in chess for 90 years with no such issue as in AoE2. New player elo is still way below average elo.
Not sure why AoE2’s system ends up with average elo = new player elo. Which common sense, a new player is not of the skill of an average player.
Possiblity they changed the original ELO formula for AoE2.
Or maybe being a 20 years old game many newcomers weren’t really newcomers. But this isn’t the case 4 years after release.
The league system like in League of Legends, also have an ELO behind it, so it’s just ELO with a flashy badge.
Chess has a different system because you get an ELO by someone judging your skill level.
You can’t do that in AoE2.
Do you have any idea what you are talking about.
ELO is a 0 sum game.
That means no points are generated out of thin air, they always come from another player.
ELO points are always moved around, never generated and never deleted.
Unless an account gets deleted I guess.
If you set the starting ELO to 500 then the average ELO per player would sink down.
The average ELO is per definition the same as the starting ELO.
There is a reason why most games don’t show the ELO rating to the players and instead hide it behind flashy rank names.
Also most games don’t use normal ELO they use advanced ELO where points decay over time and stuff like that.
You seriously think in chess there’s someone saying “this young lad shows promise, let’s put him at 1600”, not how it works. :)))
Yeah, turns out you have no idea what you’re talking about.
ELO is not a 0 sum game. There are two reasons for this. First of all, when a player plays his calibration games he wins more points than the opponent loses. In chess where you start at 1200 with the average 1500, new players usually win more than they lose, so it brings a plus overall. This is generally true for the first 30 games. In AoE2 is the opposite effect, bringing the average elo down.
Second, google the “k factor” or “development coefficient”. Long story short, the loser will not lose as much as the winner gains.
These 2 factors, making ELO, by default, not a 0 sum game.
Sure, yes, you are right, that’s why avarage ELO is bellow 1000. Also, that’s why so many are near 500 ELO, like 4k players, from 44k below 669 ELO https://www.aoe2insights.com/leaderboard/3/?page=400
We all here, and who did put the math behind ELO calculation in AOE2DE are wrong.
do you have a source for that? I don’t see a way how average elo can be anything other than average starting elo if every game is just an exchange of rating.
because in every game you just take rating away from your opponent, it’s (almost) a 0-sum game. so the only way to change the average elo is through people leaving the game
your starting elo isn’t a fixed 1000 (or any other number), but the FIDE rating depends on your performance in the first official tournament where you win or draw at least one game, and it also depends on the average rating of the players there. So if you play in a school club for years and then turn up to a tournament, beat half the people there you’ll get a pretty high rating.
no idea where you got those numbers from. FIDE ratings (international chess federation) don’t start with 1200, and there is little public data on what’s the average rating as only players above 1000 get published.
that’s not what the k-factor is. if you have a higher k-factor it means you win AND LOSE more points. it’s a multiplier to elo lost and gained. what you are saying is demonstrably false.
but the vast majority of games (chess and aoe2) is played between players of same or similar k-factor. in chess the k-factor is mostly determined by highest achieved rating (with some exceptions for children and teens), in aoe2 your k-factor is 32 (I think?) after you have finished your placement games. So in aoe2 elo is not a 0-sum game ONLY if it’s a game between a player who is playing placement matches and a player who has completed their placement matches. so it’s very close to a 0-sum game
yeah, that’s VERY close to starting elo. the difference can be explained with players who are no longer listed (no ranked games in the last 30 days) and different k-value during the first 10 (or 20?) games
again source? why are you saying chess starting elo is 1200? and which chess elo are you talking about?
Source: I used to play professional chess. International tournaments and all that.
That’s because it’s not, ELO is not designed to be a 0 sum game.
Then it must be what I said previously.
Because ELO is specifically designed for this not to happen. A new player is supposed to be “weak”, not “intermediate”. And go up from there.
A system where the new player has the same ranking as the average player is just a bad system.
Which is why I made this topic.
What you are saying is mostly true, but it doesn’t translate to what she said, which is:
None of that translates to “someone judging your skill level”. That someone is the ELO system itself, there’s no person there judging you giving his personal opinion whether you are good or bad, ELO is impartial and is not based on another player’s opinion.
Turns out they lowered the minimum from 1200 to 1000. Back in my days it used to be 1200. So what I suggest AoE2 to do, Chess already did.
soruce? if what I’m saying is demonstrably false you should have the demonstration.
K-factor is not a multiplier. It’s the maximum number of points you can win or lose in a match. If you have 2400 elo and lose to 1800 elo, you will lose less points and he will win more points. Making a difference. If you have 2400 elo and win to 1800 elo, you will win less points and he will lose less points, since normally he wouldn’t lose that many points when being beaten by 2400. Again, leading to a difference. Which results in → Not a 0 sum game.
In AoE2 yes, that’s the issue. And that’s why the problem in OP exists.
Yes, that’s the issue.
Talking about FIDE rating, the international rating. But it turns out they lowered it from 1200 to 1000. Which 1000 is even further away from the average 1500.
It stands to common sense that a new player is weaker than the average player, so having new player start with 1500 elo because the average elo is 1500 would be terrible.
