librematch:design:matchmaking:rating

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
librematch:design:matchmaking:rating [2022-10-27 22:56] – removed - external edit (Unknown date) 127.0.0.1librematch:design:matchmaking:rating [2022-10-29 03:28] (current) simonsan
Line 1: Line 1:
 +====== Rating ======
 +
 +  * Glicko2
 +  * Discussion: different Ratings and influence on Matchmaking?
 +    * 1vs1 Rating (RM/DM/EW), TG-Rating (RM/DM/EW), Arabia Rating?
 +    * or rather just a 1v1 and TG-Rating?
 +    * another possibility:
 +      * open/closed/hybrid-rating
 +      * three categories weighted
 +      * hybrid: megarandom, water, other ame modes
 +      * open could be arabia
 +      * closed something like black forest, arena, michi
 +      * if someone doesn't care and wants to play open: `!queue open 1v1,2v2`
 +    * pair Rating systems:
 +      * someone who is super strong on Arabia should be also usable on Black Forest
 +      * Similarly, islands, whoever can do Archer Micro won't go down completely.
 +      * If you consider them completely separately, you also run the risk that a player has too few games on individual ladders for decisive ratings.
 +      * it's like `core rating + open rating` and both have to adjust a bit after one game.
 +      * or `core rating + open rating + 1v1 Rating` e.g. and then you can even weight them
 +      * something like core rating counts 50% of the overall rating, open rating 20%, 1v1 rating 30%
 +      * `core rating` can be also including the starting rating from another ladder/from another rating system with a factor to include it into the core rating for migrating from other platforms (something like this is existing already on ncZone)
 +
 +AoE2DE currently does it like [[ https://www.twestaoe.net/tools/elocalculator | that ]].