Even if that were the case, changing the Elo could still screw up the victory chance calculation.
We are all just brainstorming here. The actual way would be to take all the historical data, come up with a new modal, and fine tune it with the historical data.
As somebody suggested in this thread, get new players to play like 10 games against AI. Make sure they win like 70% of those. Then, account for that in the model.
Lowering the starting elo, even if it would somewhat work in the short term, is just applying tape to fix your dilapidated house thatās leaking water everywhere.
I think most player wouldnāt notice a change in elo of <100, as thatās within normal variation for most. So the āshort termā fix, might last for several years.
We just donāt know the rate of new people trying ranked for the first time, so itās hard to tell
Iām not a MP guy (probably also because of the ELO system), but one thing which a good system should do is motivate participation, like a league system like AOE4 does it. Working towards goals is way more satisfying than losing points. Reset the ladder after a certain period of time too to make the people want to stay. Maybe add also some gimmicky rewards like profile icons from past events you can choose from once you progress to e.g. Silver league.
The past event rewards should be obtainable in some way or the other IMO.
I think resetting would be quite frustrating. In the beginning of a new season you would have lots of one-sided matchups. I think it would be more interesting to have seasons and at the end of the season you can get promoted to the next tier in a league system or something like that. but that still leaves weird situations: what if someone progresses really fast, do we keep them in lower leagues for months?
how much time though?
I linked my post where I made some calculation regarding this. Long story short: it takes 10 000 new players (at the proposed 500) for the average to change by 100 points. I donāt know at which rate we get new players, but I think this might be several years. If we set the starting elo to 800, it would even take 40 000 new players
Over time meaning after years if not a decade? @TwerDefender did the math.
Currently there are about 40.000 players in ranked. If the new starting elo is 500, if there are 40.000 new players, the average elo would be decreased to 750.
How many years it would take for the pool of ranked players to double? And thatās not even āthe average Elo will settle around the starting Eloā, itās just reaching the middle of the road. With 80.000 new players (12.000 players in total) the average elo would be 625. So you still wouldnāt reach the starting elo with the average.
By the time the average Elo will settle around the starting Elo, I will consider pensions plan, if AoE2 is still around and the computer didnāt become obsolete, so itās only short term in the lifespan of a greenland shark.
