vijasi's blog

Is Rating Inflation Real?
By vijasi, history, 4 years ago, In English

With increased number of participants nowadays rating seems to increase easier than before.

Have titles like "candidate master", "master" etc. lost their original significance and meaning?

What are your thoughts about the same?

  • Vote: I like it
  • +36
  • Vote: I do not like it

Comments Comments (19)
| Write comment?
»
4 years ago, # |
  Vote: I like it +14 Vote: I do not like it

Agree

→ Reply
»
4 years ago, # |
← Rev. 4   Vote: I like it -24 Vote: I do not like it

i think its time to do some upate with colors. for example :

Black : 3200 — inf

Red : 2800 — 3200

Orange : 2300 — 2800

Purple : 2000 — 2300

Blue : 1800 — 2000

Cyan : 1500 — 1800

Green : 1200 — 1500

Grey :1200 — 0

what do u think about that?

→ Reply
  • »
    »
    4 years ago, # ^ |
      Vote: I like it +10 Vote: I do not like it

    I think changing colors randomly when it just feels right is stupid. Either you have to consistently change colors to match with user ranking ratios or you never change them, and I'm pretty sure most people don't want them to constantly change.

    → Reply
    • »
      »
      »
      4 years ago, # ^ |
        Vote: I like it +20 Vote: I do not like it

      The first thesis is pretty strange. What's the problem? It's quite logical to change something if you feel it looks wrong right now.

      → Reply
      • »
        »
        »
        »
        4 years ago, # ^ |
          Vote: I like it -7 Vote: I do not like it

        Perhaps I'm in the minority in thinking this, though I do not think I am, but for me I want to set goals and have comparisons that are always calculated in the same way, whether that's through ratios or static barriers or something else similar. If the bounds change arbitrarily at random whenever it just feels like it should, you have no ability accurately reflect on what the colors mean. However through consistency, people get a reasonable feeling of what each color means so they seem like good goals and comparisons.

        → Reply
        • »
          »
          »
          »
          »
          4 years ago, # ^ |
          ← Rev. 2   Vote: I like it +1 Vote: I do not like it

          We don't say it's going to be changed every month, but the latest time it's been (except some cosmetic changes) was $$$5$$$ years ago. I think one time per $$$5$$$ years would be rare enough to avoid this :)

          Also, I can say that the recent time I am checking my progress not using rating as a metric, but only a place at the top list, because probably inflation has become very fast now(but it's only a subjective view and should be investigated).

          → Reply
          • »
            »
            »
            »
            »
            »
            4 years ago, # ^ |
              Vote: I like it +10 Vote: I do not like it

            I will agree that if it is changed consistently on some founded basis, then that would fit exactly the kind of idea I would like. For example if the colors were reformatted once a year based on amount of people that should be in each color, not arbitrarily, that would be a great way to keep the colors having the same meaning overtime and help reduce rating inflation.

            → Reply
  • »
    »
    4 years ago, # ^ |
      Vote: I like it +27 Vote: I do not like it

    2300 — purple, seriously? It can be checked, but I don't think that the ratio red_users/all_users is much bigger than it was e.g. one year ago.

    → Reply
    • »
      »
      »
      4 years ago, # ^ |
        Vote: I like it +45 Vote: I do not like it

      It is not obvious to me if red_users/all_users is in fact the right metric to go for.

      If I ask my grandma to register at Codeforces, will it make my skills more impressive?

      → Reply
      • »
        »
        »
        »
        4 years ago, # ^ |
          Vote: I like it -10 Vote: I do not like it

        I would argue that just the number of red users should be a good metric and that is obviously increasing.

        → Reply
        • »
          »
          »
          »
          »
          4 years ago, # ^ |
          ← Rev. 2   Vote: I like it +16 Vote: I do not like it

          I think there is some sort of balance point in the middle. Both extremes are less than perfect. On the one hand, being top 100 in the world right now is likely harder than it was 10 years ago, because there are more people doing it more seriously and over longer time span. On the other hand, being in top 1% in the world is likely easier now, because "I've been told that I need to do this to get a job" / "I will try this out of curiosity now when there are so many cool sites with cool competitions" kind of growth is likely even faster than the growth in number of people who take it seriously.

          → Reply
          • »
            »
            »
            »
            »
            »
            4 years ago, # ^ |
              Vote: I like it -13 Vote: I do not like it

            In my opinion division should really be division. No overlapping contest and when you enter Div 1. your rating is "reset" to a fixed value and then only Div 1. contests count.

            → Reply
»
4 years ago, # |
  Vote: I like it +86 Vote: I do not like it

How can rating inflation be real if CF isn't real?

→ Reply
  • »
    »
    4 years ago, # ^ |
      Vote: I like it -18 Vote: I do not like it

    How can you be real if CF isn't real? xD

    → Reply
»
4 years ago, # |
  Vote: I like it +70 Vote: I do not like it

What do you mean? My rating has not increased.

→ Reply
  • »
    »
    4 years ago, # ^ |
      Vote: I like it +4 Vote: I do not like it

    I guess it's prevalent more for Div 2 as Div 2 participants have increased more.

    → Reply
»
4 years ago, # |
  Vote: I like it +2 Vote: I do not like it

But your last contest was almost a year ago!

→ Reply
  • »
    »
    4 years ago, # ^ |
      Vote: I like it 0 Vote: I do not like it

    He can compare his old results with the current results of cyans

    → Reply
»
4 years ago, # |
  Vote: I like it +3 Vote: I do not like it

I remember there was a lot of concern that ratings were getting deflated not that long ago. Someone discovered in a contest the total number of rating points gained by all contestants was less than the total number of rating points lost.

→ Reply