Please turn on javascript in your browser to play chess.
Only Chess Forum

Only Chess Forum

  1. 01 Oct '16 18:56
    I have a question regarding Fritz (15) and the playing strength slider for Rated Games. No matter which engine I install (from a very weak 1500 Tarrasch to a very strong Stockfish e.g.), the slider always has a range between 1039 to 2439 (give or take).

    Can someone explain what's going on here?
  2. Subscriber Paul Leggett
    Chess Librarian
    03 Oct '16 00:19
    Originally posted by peewee678
    I have a question regarding Fritz (15) and the playing strength slider for Rated Games. No matter which engine I install (from a very weak 1500 Tarrasch to a very strong Stockfish e.g.), the slider always has a range between 1039 to 2439 (give or take).

    Can someone explain what's going on here?
    Since the program has no way to actually or reasonably assess a relative ELO rating for whatevery engine you use in conjunction with whatever hardware you are using, it is probable that the 1039-2439 range is just the standard range for the sliding adjustment.

    I would suspect that the system uses some kind of handicapping mechanism such as how many ply forward it can look which would be independent of the engine algorithm itself.
  3. 03 Oct '16 09:03
    Thanks Paul. Do you happen to know if Fritz Elo rating follows the possible Elo ratings of the engine? What I mean is this: some engines use a playing strength (like Stockfish's 1-20) but some engines do indeed try to estimate an Elo rating (like Hiarcs or Delfi).

    So: does Fritz "follow" the engine's Elo *if possible* (when using Hiarcs, Delfi, etc.) or does Fritz always use it's own method to dumb down the engine?
  4. Subscriber Paul Leggett
    Chess Librarian
    03 Oct '16 22:14
    Originally posted by peewee678
    Thanks Paul. Do you happen to know if Fritz Elo rating follows the possible Elo ratings of the engine? What I mean is this: some engines use a playing strength (like Stockfish's 1-20) but some engines do indeed try to estimate an Elo rating (like Hiarcs or Delfi).

    So: does Fritz "follow" the engine's Elo *if possible* (when using Hiarcs, Delfi, etc.) or does Fritz always use it's own method to dumb down the engine?
    The short answer is "no". The following link touches on the subject, and is also a very good exposition about ratings in general, including internet ratings:

    https://en.chessbase.com/post/fritz-ratings-vs-
  5. 04 Oct '16 06:39
    Thanks a lot Paul. Very informative.