Elo-like rating interpolation?

Question

Analytics

Hey Habr!
There was such a need. There is some sports community, abstractly speaking, lovers of cymbal hitting. Competitions are regularly held in this case and a certain Elo-like rating is built (I can’t say more in detail - I don’t know the exact calculation method myself), expressed as a single number.
The task is as follows: based on the results of the selected person, I want to be able to calculate his expected rating in a month, six months, a year. Can someone show in which direction to dig? Maybe someone has already had to implement this (after all, surely there are services for predicting their rating after a certain period of time for chess players?). Can you advise the Talmud, in which the systems of such a rating calculation are generally described in a relatively clear and understandable way?
The task comes down to some kind of interpolation, which should be quite realistic - for example, well, even in a hundred years, no cymbal smasher will be able to reach a rating of 5000, when the current world record is about three thousand.
Thanks in advance.

Answer 1

Wikipedia is well written about the calculation system. The Elo rating allows you to get the expected value of the number of points that plate breaker A will receive in a face-to-face confrontation with plate breaker B. If you know, for example, which matches (and with whom) each player is facing, then you can calculate the expected value of the change in his rating for each player in these matches. For example: “Tomorrow A plays with B. Judging by the rating, from will score 10 points in this match, and his rating will increase. The day after tomorrow he plays with B. Let's take into account the change in rating A from the previous match, and see what he can count on this time”, and so on.
Another thing is that breaking plates and chess are two different things. For example, if you're hitting cymbals, you (probably) need some fitness. In addition, you can get injured in the process, which will affect the next few matches (if in your case all this, of course, applies). In sports analytics, all of these things are also taken into account, as well as the history of meetings with a particular opponent, the success of a performance on a certain platform for breaking plates, or, say, the number of wins and losses in breaking a certain brand of plates.