Where was we intuitvely know that a play who just started playing AoE2 ranked isn’t an average player, but a new player. Meaning new players in ranked are going to have a really really hard time.
In my opinion, the starting elo for new players should be dropped to 700. But this doesn’t mean everyone gets -300 elo. Everyone stays the same and new starting elo is 700. This, or every existing player gets +300 elo and the starting elo remains the same, 1000. It’s the same thing.
You clearly start at the Beginner-Novice level but not the lowest elo possible. You don’t start at 1500 elo.
While it’s reasonable to assume that ELO is a 0 sum game and 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. Thus, if you set the starting ELO to 700 then the average ELO per player would eventually sink down to 700.
In practice, it would take decades before the average elo drops to 700 elo. AoE2 currently has about 40.000 ranked players in total. It would requre another 40.000 players to drop the average elo to 850 and 80,000 new players to drop the average elo to 800. So by the time AoE2 has x3 more players than it has today, the average elo will still not reach the starting elo.
This is something HERA talked about:
His idea is even more drastic, starting elo: 500.
But at the same time new people are getting better, so…
I would argue that different medals should give different ELO points: Maybe 15, 20 and 25 for bronze, silver and gold. Numbers pulled from my onager, but still, the points stands.
For others who watch, the point in this thread starts being discussed at 16:00mins into the video.
But I agree. As a relatively new player, playing since mid Jan 2025 with just 98 games (36 wins), I’ve now semi-stabilised at under 400 elo. I can somewhat understand why new players don’t start at 0 elo too as, in the post I recently created on this forum, climbing that elo ladder is so tough. Now, even decent players will take so long to rise up the ranks as even at my elo, you see the tactics of the pros being used i.e. meta gameplay. So the lack of progress could also make people feel daunted and quit that way.
That is why I would argue that dropping the starting elo (this thread) and seasonal elo resets (my thread) should be considered.
I don’t have an opinion on the suggestion but I’m interested in the maths behind this claim:
If the starting elo is 1200 and it’s a zero-sum game, how come the mean elo isn’t 1200? I think I must be misunderstanding what is meant by “zero-sum” here…
Thanks! So this shows that the starting elo problem on chess.com is actually much worse than that in AoE2 – the starting elo is at the level of the top 10% of players and the median is less than half that. Looks like the OP has been misled by a bad graph and chess’s snobbish ranking terminology, where you have to be in the top ~5% to be considered “weak intermediate”.
If you are talking about the FIDE Elo, the starting Elo is determined by your performance in the first official tournament.
On Chess[.]com it is 1000, on Lichess it is 1500. But both values are the average Elo of the playes in the platform. Although they are different values, both indicate a player of the same strength.
Choosing a starting value that is not the average strenght of the players creates other problems in the long run.
I think it depends on the fact that Chess.com shows inactive accounts, whereas in AoE2 they disappear from the ranking. So you don’t know how many players have played a few games and then left multiplayer alone.
