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

Only Chess Forum

  1. 06 Nov '16 12:46
    They have done a computer match up v the top players.

    Carlsen comes in at 227, Karjakin 202 , Fischer 336 & Kasparov 498.

    The top is Wesley So with 89.32%


    http://chess-db.com/public/top100pqi.jsp
  2. Subscriber Ragwort
    Ex Duris Gloria
    07 Nov '16 08:58
    I see the list has Capablanca at number 2 with 92%. He must have had a super charged abacus to train against! I followed some links from the preamble to the chess life article on professor Regan's anti cheat tests. Fascinating, by matching computer human choices in specific complex positions he can give a likelihood of a human of a specific grade finding a move. This gives mathematical credence to something we have all felt for a long time - you can tell if you're playing an engine .... or Capablanca with an abacus.
  3. 07 Nov '16 11:15
    Hi Ragowrt,

    and yet Capa's high score....we know for a fact he was not using a computer.
  4. Subscriber moonbus
    Uber-Nerd
    08 Nov '16 07:07
    Perhaps the match-up indicates that some players (such as Xu, Gheorghiu, and Capablanca) tend to think in highly schematic ways, which is how computer programs work, whereas other players (Fischer, Kasparov) are highly idiosyncratic in their thinking, which is much harder to code in a finite set of unambiguous algorithms.
  5. Subscriber Ragwort
    Ex Duris Gloria
    08 Nov '16 08:17
    Originally posted by moonbus
    Perhaps the match-up indicates that some players (such as Xu, Gheorghiu, and Capablanca) tend to think in highly schematic ways, which is how computer programs work, whereas other players (Fischer, Kasparov) are highly idiosyncratic in their thinking, which is much harder to code in a finite set of unambiguous algorithms.
    Some have said that Capa tended to go for simpler positions and in that sense may have given himself less chance to go wrong. When the quality of play index first came out comparing world champions it was the "positional" players who scored highest. My suspicion is that in order to maintain positional control great tactical vision is necessary.
  6. Subscriber moonbus
    Uber-Nerd
    08 Nov '16 10:48
    Originally posted by Ragwort
    Some have said that Capa tended to go for simpler positions and in that sense may have given himself less chance to go wrong. When the quality of play index first came out comparing world champions it was the "positional" players who scored highest. My suspicion is that in order to maintain positional control great tactical vision is necessary.
    Hmmmm. Makes me think. What computer programs excel at is calculating variations, once the search tree has been pruned by general principles. Reti was once asked how many moves in advance he routinely calculated, and famously answered "none". Of course, that was hyperbole; he was certainly capable of calculating when he had to. But his answer illustrates a good point about the difference between human- and computer-play. So this suggests that the humans who match-up closely with computer-play may be employing a similar thinking pattern to those implemented by computer programmers, whereas the Reti-ists among us tend to think rather more abstractly and less calculatively.