Overview
There are 1409603 rating changes in 792 rated contests (by 2018/07/14, including those mentioned in the "Something Weird" section). There are 142117 users with at least one rating change. The overall average of deltas is -6.671139.
Rating Changes Distribution
(Well, it looks like a boa constrictor eating an elephant...)
Top 20 Events: Positive Rating Change
Top 20 Events: Negative Rating Change
Top 10 Users: Consecutive Positive Rating Changes
| # | Who | # consecutive |
|---|---|---|
| 1 | tourist | 21 |
| 2 | stormtrooper1859 | 15 |
| 3 | RAD | 14 |
| 4 | xyz111 | 14 |
| 5 | guoyu1098 | 13 |
| 6 | KK_ndv | 13 |
| 7 | The_Unbeatable | 13 |
| 8 | Saiphani724 | 12 |
| 9 | harshit_ag456 | 12 |
| 10 | vepifanov | 12 |
| 10 | Matei.Chirita | 12 |
Top 10 Users: Consecutive Negative Rating Changes
| # | Who | # consecutive |
|---|---|---|
| 1 | worse | 31 |
| 2 | Mhammad1 | 25 |
| 3 | ahmed_25 | 22 |
| 4 | newbie314159 | 21 |
| 5 | NelsonMondialu | 19 |
| 6 | .o. | 18 |
| 7 | Peregrine_Falcon | 18 |
| 8 | Adai_is | 17 |
| 9 | THton | 17 |
| 10 | MisterMax | 16 |
Top 10 Contests: Positive Averaged Rating Change
| # | Contest | Avg. Δ | # participants |
|---|---|---|---|
| 1 | Manthan 2011 | 34.309 | 745 |
| 2 | Codeforces Beta Round 88 | 21.053 | 1103 |
| 3 | Codeforces Beta Round 81 | 20.353 | 1034 |
| 4 | Codeforces Beta Round 87 (Div. 1 Only) | 16.067 | 416 |
| 5 | Codeforces Beta Round 91 (Div. 1 Only) | 15.448 | 420 |
| 6 | Codeforces Beta Round 92 (Div. 1 Only) | 15.140 | 365 |
| 7 | Codeforces Beta Round 85 (Div. 1 Only) | 14.957 | 398 |
| 8 | Codeforces Beta Round 83 (Div. 1 Only) | 14.909 | 351 |
| 9 | Codeforces Beta Round 86 (Div. 1 Only) | 14.389 | 355 |
| 10 | School Regional Team Contest, Saratov, 2011 | 14.331 | 580 |
Top 10 Contests: Negative Averaged Rating Change
| # | Contest | Avg. Δ | # participants |
|---|---|---|---|
| 1 | VK Cup 2012 Round 3 (Unofficial Div. 2 Edition) | -31.142 | 643 |
| 2 | VK Cup 2012 Round 2 (Unofficial Div. 2 Edition) | -28.747 | 707 |
| 3 | Croc Champ 2012 - Round 2 (Unofficial Div. 2 Edition) | -27.681 | 627 |
| 4 | Codeforces Round 112 (Div. 2) | -27.462 | 1518 |
| 5 | Codeforces Round 113 (Div. 2) | -26.989 | 1443 |
| 6 | Codeforces Round 324 (Div. 2) | -20.562 | 4333 |
| 7 | Croc Champ 2012 - Round 2 | -20.151 | 338 |
| 8 | Codeforces Round 114 (Div. 2) | -19.963 | 1395 |
| 9 | Codeforces Round 326 (Div. 2) | -18.719 | 3838 |
| 10 | Codeforces Round 323 (Div. 2) | -18.305 | 4728 |
Rating: Sum, Average, Standard Deviation and Skewness
A user is included in the sum only after their first participation in a contest.
Note that we can observe changes of rating system (November 2011 and October 2015) on the graph.
(I wondered if there were changes in March 2012, since the top 5 contests with most negative average delta are all in March or April 2012. However, I didn't find any related posts.)
Something Weird
- In Codeforces Round #FF (Div. 2), kasim appeared in both #1095 and #1986.
- In Codeforces Round 270, yuki2006 appeared in both #2876 and #3520; a00920 appeared in both #1097 and #3602.
- In Codeforces Round 338 (Div. 2), bohuss appeared in both #256 and #3477; InnocentFool appeared in both #555 and #3477; elgris appeared in both #884 and #3477; Altitude appeared in both #2439 and #3477; mohamedazab appeared in both #2831 and #3477. (Such a buggy round?)
- In MemSQL Start[c]UP 2.0 - Round 2, FreezingCool had a rating change of +5, but it wasn't displayed on the user page. (Can somebody explain why max rating is smaller than contest rating?)
- In Технокубок 2018 - Отборочный Раунд 4, Roma_Pussy had a rating change of -83. However, it seems that the rating is not actually changed. After that, his ratings displayed on "contest rating" and "rating change" in contest pages was 83 higher than those on the user history graph.
- In contest pages, ratings less than 100 are displayed as 100. 8 users have been affected: worse, MisterMax, newbie314159, errorerror, Mhammad1, KHOKHO_IS_N00B, Peregrine_Falcon and NIKABB_IS_N00B_LIKE_ME.
- https://codeforces.net/contests/1004 (Codeforces Round 495 (Div. 2)) is redirected to https://codeforces.net/contests/1002 (Microsoft Q# Coding Contest - Summer 2018).
Very interesting stuff. I thought I heard codeforces was undergoing rating inflation but the last plot seems to show the opposite...?
It is possible that the mean lowers while the standard deviation gets higher; which can lead to something similar to rating inflation.
I've added graphs of standard deviation. It didn't change too much in recent years.
Very interesting, thanks!!!
Auto comment: topic has been updated by adrien1018 (previous revision, new revision, compare).
What the hell happended with King_George ? From international grandmaster to pupil. Some sort of a rage quit, I guess.
Probably just wants to have a fancy graph like .o.
Intentional feeding
Interesting. How did you get Something Weird information?
I didn't intend to get that information. However, when I was arranging all records, some assertions in my program failed. After some checking, I realized that there's something wrong with the data. Since it was necessary to correct them one by one, I wrote the section conveniently.
[DELETED]
I think you can update 3rd position of Positive Rating Change. It's mine now. +670
Auto comment: topic has been updated by adrien1018 (previous revision, new revision, compare).