Every player starts out with e^1 skill points. Rating is calculated backwards in time from the previous full day. Time Deviation of each player is calculated for each player based on how quickly their performance is changing over time as the rating is calculated beckwards in time, and for fluctuations that can't be explained by time deviation, K factor as it's usually applied is pretty silly. It should be a continuous scale based on previous performance, as it is in Mamumamu0413's, not semi-arbitrary values based on rating points, and I can only think of narcissism as an answer for why FIDE would do that. Their rating system is objectively worse than Mamumamu0413's. Elom ratings follow this logic, and a players K-factor is based on the range of ratings they beat and lost to, so if to players have the same rating, the one who got that rating winning against higher ranked and losing against lower ranked players would have the higher Rating Deviation/K-factor, like Mamumamu0413 I would use Rating Deviation which is better than K-factor.

What's mine? Can I have one.

Haha, yes, having AGA and EGF results in the system along with those of the top pros is the intention!

Elom ratings will begin 2024 March, results taken into account for performance ratings are from open international tournaments in the preceding 16876 days (12 years), national tournaments and female, youth or senior international tournaments in the preceding 1443 days (4 years), female, youth or senior national tournaments in the preceding 28 days. Ratings are calibrated by comparing performance ratings from previous calibration cycles to produce a final rating. As the win and loss results for any period can be updated at anytime, this allows for future correction, essentially making this a whole history rating system as ratings are calculated in reverse chronological order to which game are played . . O