Is the rating predictor not working correctly or is the rating assigned wrong? For some, it deviates a lot in the negative direction, for some a lot in the positive direction and for some, it approximately matches with the predictor. Can someone please explain why is it occurring? In the previous contests, it had just the error of +- 5, now it's deviating a lot in both directions.
MikeMirzayanov can you please give any update, if it's a mistake from your side or not? These deviations I am seeing for the first time, and even some months back, there was a mistake in assigning the ratings.
→ Обратите внимание
→ Трансляции
→ Лидеры (рейтинг)
| № | Пользователь | Рейтинг |
|---|---|---|
| 1 | tourist | 4009 |
| 2 | jiangly | 3823 |
| 3 | Benq | 3738 |
| 4 | Radewoosh | 3633 |
| 5 | jqdai0815 | 3620 |
| 6 | orzdevinwang | 3529 |
| 7 | ecnerwala | 3446 |
| 8 | Um_nik | 3396 |
| 9 | ksun48 | 3390 |
| 10 | gamegame | 3386 |
| Страны | Города | Организации | Всё → |
→ Лидеры (вклад)
| № | Пользователь | Вклад |
|---|---|---|
| 1 | cry | 164 |
| 1 | maomao90 | 164 |
| 3 | Um_nik | 163 |
| 4 | atcoder_official | 160 |
| 4 | adamant | 160 |
| 6 | -is-this-fft- | 158 |
| 7 | awoo | 157 |
| 8 | TheScrasse | 154 |
| 8 | Dominater069 | 154 |
| 8 | nor | 154 |
→ Найти пользователя
→ Прямой эфир
Codeforces (c) Copyright 2010-2024 Михаил Мирзаянов
Соревнования по программированию 2.0
Время на сервере: 14.11.2024 19:23:35 (f1).
Десктопная версия, переключиться на мобильную.
При поддержке
Carrot had a pretty good prediction compared to CF-Predictor for me.
I use both, you can compare from the previous rounds, CF-Predictor almost predicts correct rating with an error of +-2, while Carrot has more error. In this contest only, rating changes are deviating a lot and even it's not the same for all the users. Just compare the ranks 96 in Div2, there are two users, for ChenHuaXin, rating predictor: +154 (actual: +156) and for YL-9, rating predictor: +106 (actual: +94). There are more such examples, generally what happens, if the rating predictor says +10 (let) and the actual change is +12 (let), then that 1-3 difference will be for all the users.
I think there is some issue with the rating change this time, as it happened some months back.
Can you explain how to use CF-rating predictor ? idk how to use it
To use it you should be doing CF on a Google Chrome. If you do so: https://chrome.google.com/webstore/detail/cf-predictor/ocfloejijfhhkkdmheodbaanephbnfhn.
There you install said extension and nothing more needs to be done. After successful installation you will see an approximate change in points for anybody you want in a "standings" section for each contest (if you follow the above link you'll see a picture with an example of how it will look like, the rightmost (delta) column is responsible for showing an approximate rating change).
My mistake, I knew about this extension, actually I wanted to know how to use virtual rating predictor from this website : https://cfviz.netlify.app/virtual-rating-change.html
And how much accurate is this virtual rating predictor ?
For me for the last contest it is 20 points difference between a prediction and real points change. As for me that inaccurate at all.
To check it yourself: in the first "contest ID" field you put a number that you can see at the end of URL for a particular contest. Then you put your points from standings from a "=" column in a "points" field. After you put your CF name (handle), amount of penalties you had in a contest (0 if none). And then you fill the last section with your previous (old) rating.