Effects of changing the starting elo to 500

I calculated this to respond to a reddit post and thought it would be interesting for the people here as well. I used to be quite opposed to lower the starting elo, as it would just pull everyone else down to the new average, but I failed to take into account how long this would take.

atm we have about 40 000 active players. I don’t know at what rate new players join so let’s do the calculations for the new average elo depending on how many people join (assuming the current average is 1000): (EDIT: new starting elo in this caluclation is 500)

after 1000 new players, the average elo would be around 988
after 5000 new players, the average elo would be around 935
after 10000 new players the average elo would be around 900
after 20000 new players the average elo would be around 833
after 40000 new players, the average elo would be 750
that difference in elo corresponds to about .75, 3.5, 6.25 and 10.5, 15.6 extra losses (average elo change in a game is 16).

So the question you need to ask yourself is “Am I ok to have 6 extra losses in the time it takes for 10000 new players to join the ranked ladder so that they have a better experience

It would be really interesting if a project manager or someone with this knowledge could enlighten us on how many people play ranked for the first time in a given time interval.

Ideally we would have a system that somehow determines the starting elo based on previous experience (even just a self-assessed ‘i can beat hard AI’)

4 Likes

500 starting elo is too low. We need to protect really low elo players from average new players and smurfs. 750 starting elo is more reasonable.

I picked a low number to see how it affects average rating. if a higher starting elo is picked the convergance to it is even slower. I don’t have a strong opinion for any number in particular. I think something like 800 with a K-value of 64 until the first win and then a K-value of 32 (as it is now) could be good

2 Likes