Which brackets would this apply to? I don’t think it fully applies to the 850- ELO bracket, and that’s the lowest ELO bracket on AoEstats. Every skill you mentioned is a skill I can use, and I’m only ~1000 ELO. I’m fairly certain I’ve seen this sort of argument used to discount data from the 1200+ bracket, and it just doesn’t hold there.
But frankly, the stats show that civ differences are about as meaningful for low ELOs as they are for high ELOs. How meaningful each specific bonus is may change, but the overall variation from civ differences is still approximately the same.
yes but many players quit ranked gaming since 1000 is a very high starting elo and this means every new player would lose most of their initial games until they match with other new players. And losing 100 or 200 elo is going to have a substantial negative impact than losing 20/30 elo.
For about a decade this is how balancing has been done. Militia line, infantry uu like Jaguars, Samurais and Teutonic knights were always popular in low elo games but they used to be very weak in mid and high elo games. After the buffs, low elo games have remained as they are while the units are slightly more usable in competitive games. So unlike what you believe, making balance changes based off higher elo games isn’t going to negatively impact low elo games. In general balance changes are going to have very low impact on low elo games because the civs and bonuses are almost never used the way they are supposed to be used in such games.
It takes a very long time for the starting elo to become the new average elo (technically forever).
For the average to go just half way from an old stating elo to a new starting means replacing half the playerbase with new players, or doubling the playerbase (or some mix of the two).
To give an example: assuming we have 40 000 1v1 players, current starting elo is 1000, new starting elo is 800.
For the average to move to 900 requires:
40 000 new players
or
20 000 old players to leave and be replaced by 20 000 new players
or a mix. for example:
10 000 old players leave and 30 000 new players join.
a different (very informative) thread in this forum had some data on number of people trying out 1v1 for the first time each week. Based on those numbers it would take several years for this shift to happen. a 100 elo shift over several years would not be noticeable.
Simply changing the number of the starting elo wouldn’t fix the problem. It would just make 800 (or what ever number you pick) the new 1000 over time. The inherent problem is that most new ranked players are obviously going to be worse then the average long time ranked player.
One idea I had is giving people placement matches vs. the AI but that wouldn’t fully fix the problem of shifting the elo down because then new players would get a below 1000 rating on average. Also they would sometimes have to make the AI play against ranked players to figure out what elo the AI is on. And I’m afraid that the extreme AI might be below 1000 elo.
Not sure what you mean with low elo here because they were only popular with very casual players, like super low elo.
You need to remember that even an 800 elo player is probably a lot more experienced then the average none ranked player.
I know that it’s technically forever until exactly 800 would be the new starting elo but the skill level at the 800 mark would constantly rise from the point you make it the starting elo, so eventually you have to lower it again, or alternatively constantly lower the starting elo which could get out of hand quickly.
So it would always just be a temporary solution.
As I wrote earlier in the post. Ranking matches vs. AI could help with getting people into the right elo. And of course better tutorials could help people just being better when they start playing ranked.
Maybe they should hire some pros to make some more advanced tutorial missions which give you insights into playing each of the civilisations. Like making Chinese start tutorial that helps people train how to master their unique start by giving you medals depending on how well you have done. I’m not the first one to suggest this.
What you’re saying is not the point of my discussion, its just to imply that a game with high learning curve will have left shifted elo distribution. So even though most people are in the lower elos it includes people who have quit, are new, smurfing etc and the number of active true 1k elo players might be lower. So the argument that most people are less than 1k doesn’t imply that the stats from them is informative.
Anyways I do understand the system theoretically. But it doesn’t take psychology into account especially younger people’s. And I’m also not talking about resetting all player elos just re-adjust those who are less than 1100 and have played less than 20 games. Those who have played many games will retain their current elo while newcomers won’t have to get stomped. Losing 2 or 3 games, 50-60 elo is going to be significantly less demotivating than losing 300 elo by getting crushed several games in a row. Eventually there has to be some sort of elo determination test from Art of war (with more scenarios).
And even if you fix this issue, my point is balance changes shouldn’t be made based on casual games. That will imply nerfing weak, poorly designed units like Teutonic knights or Battle Elephants and buffing CA, Mangudai which won’t make any sense. Balance changes are for competitive games where both players know how to play their civ, the map, the match-up.