2013 Candidates: What does the tiebreak system really mean?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • 2013 Candidates: What does the tiebreak system really mean?

    On various chess forums there has been a lot of heated opinion on either side of the 2013 Candidates' tiebreak system, which has determined that Carlsen will play Anand for the World Championship later this year. Although in the chess point system 1 win = 2 draws, in the 2013 Candidates' it was agreed beforehand that 1 win (along with a loss) is greater than 2 draws when it comes to breaking ties in points.

    Many people would prefer that some kind of tiebreak match be played between Carlsen and Kramnik. Others feel it is right that Carlsen advances. Still others feel Kramnik should advance, with no further tiebreak matches played.

    There is little doubt that standard organized chess is at a critical juncture in history. These 2013 Candidates have provided some of the best top-level chess entertainment in years, helped by the internet coverage. Human vs. human elite-level chess has rarely seen greater drama (albeit appreciated only by serious chess players). Even the implaccable Carlsen has been exposed as somewhat weaker than expected -- and not so implaccable.

    The chess establishement -- FIDE, that is -- has decided that wins (and losses!) supercede draws in determining which of two players is greater.

    For many years now, draws have been the bane of standard chess at the highest levels. If chess has failed to grab international general public attention or to attract sponsors in droves, excessive draws are pointed to by some as being to blame, and that goes beyond the much-criticized non-fighting "GM draws". It includes even fighting draws.

    What then does this really mean at this critical juncture for chess? Obviously it is an attempt to drive the top players to fight for wins harder than ever before. Kramnik has just lost approximately $1 million because it is his style to play "safer" than is deemed good for chess.

    (An aside: it is ironic that it is Kramnik who is robbed by this tiebreak of the chance to play for the WC. Kramnik was the one who not so long ago complained in an interview piece that it is getting harder and harder for the elite GMs to gain even the slightest advantage over the board, and that chess was becoming far too focused on opening preparation. He commented that some sort of change is needed. Perhaps this result will see Kramnik spearheading an effort to bring about such a change.)

    Is this the beginning of a sea change for top-level chess in which draws will eventually be not just frowned upon, but eradicated altogether? And if so, it that a good thing? Is it really that the secret to growing chess popularity worldwide is to simply make it "fighting chess" in the supreme sense, meaning that draws are simply not possible?

    My personal opinion is that more steps than just eliminating draws are needed, but... I've made contributions here towards draw reduction with 2 separate ideas (although falling short of outright banning of draws, either idea would make draws very very rare):

    http://www.chesstalk.info/forum/showthread.php?t=7398

    http://www.chesstalk.info/forum/showthread.php?t=8462

    The first would basically turn chess into a Shogi-like variant only in drawish circumstances, to keep play going and making wins possible where they weren't before.

    The second would take computer engines out of chess altogether and for many years to come. Of the two, I prefer this second method. The other thing this second method would do is to allow playing chess for money online possible. No chess engine would be of any help to anyone in online play, and it would take years for engine authors to achieve the level of AI needed to make decent decisions under this system.

    (For those of you who have been following my posts over the past year or so, you will recall that I am working on a chess variant that will be playable for real money online, and that I will make it available (to a select few initially) before this year is out. It will have a chess-playing component, but the fact that this will be played online for real money means we have to handle the issue of chess engines. Thus my second proposal above, which will constitute the chess playing component of the game... although the first idea above might also be added. Final decision has not been made yet.)

    The true question seems to be: is it enough in this 21st century for chess to be decided merely on the ability of players to choose the best move more often than anyone else? We know that computers can already outperform humans in this regard. However, one thing computers can't do is play like Chucky, which is to say, they can't surprise. Either of my suggestions given above would allow for even more surprises than Chucky provided in the 2013 Candidates.

    If what we want in top level chess is exciting drama leading to clear-cut winners and losers (almost always), either of my suggestions would seem to be at least equal to what has decided the winner of these 2013 Candidates.
    Only the rushing is heard...
    Onward flies the bird.

  • #2
    Re : 2013 Candidates: What does the tiebreak system really mean?

    Originally posted by Paul Bonham View Post
    I am working on a chess variant that will be playable for real money online, and that I will make it available (to a select few initially) before this year is out.
    To a select few only? This is the first time you write this, I think. Why would you want to do that?

    Comment


    • #3
      Re: Re : 2013 Candidates: What does the tiebreak system really mean?

      Originally posted by Louis Morin View Post
      To a select few only? This is the first time you write this, I think. Why would you want to do that?
      Louis, I don't know how much you may know about software development, so forgive me if I explain something you may already know. When a software product is ready for overall functional testing, the first stage is called "Alpha" test. This is done internally, by the entire software development team and project managers. Bugs and functional shortcomings are discovered, written up, and addressed.

      The next stage is called "Beta" testing. This involves the software being distributed to a very limited audience external to the company doing the development. People sign up to be Beta testers, and the difference from Alpha testing is that these people are true users, seeing the software for the first time. Thus, Beta testing discovers how the software holds up to being used in the real world, because the internal people have "prejudices" (for lack of a better term) in that they know not to do certain things at certain times and so on, whereas real external users will do all the "dumb" things that Alpha testers avoid doing. Beta testing is thus more complete and realistic than Alpha testing.

      I still don't have a definite time frame, but before this year is out we will sign up Beta testers and they will be the select few I mentioned. I'm still hoping this can be underway by summer, with a fall date for launch of the web site. But there are funding needs that will decide the timeline. I would definitely like for you to be among the Beta testers, as you have been at least tolerant of my ideas. :)
      Only the rushing is heard...
      Onward flies the bird.

      Comment


      • #4
        Re: 2013 Candidates: What does the tiebreak system really mean?

        Originally posted by Paul Bonham View Post
        On various chess forums there has been a lot of heated opinion on either side of the 2013 Candidates' tiebreak system
        I haven't read other forums, but like the tiebreak system used in London. Any tiebreak system is essentially artificial -- when two or more players tie for first, none stood out so much to deserve victory. If you don't like tiebreaks - prove that you're the best, and win outright. Rapid or blitz tiebreaks, while letting the players decide, are at a different time control than the event itself. The system used in the candidates meant that the result was based on normal time control games only. A substantial benefit of the candidates tiebreak rules is that it sets up situations like we just saw in the 14th round - where both Carlsen and Kramnik had to play for a win - even though they were leading the event. If a rapid playoff was on the cards - I expect there would have been a lot less risk-taking / unexciting draws instead.

        Comment


        • #5
          Re: 2013 Candidates: What does the tiebreak system really mean?

          Originally posted by Walter De Jong View Post
          I haven't read other forums, but like the tiebreak system used in London. Any tiebreak system is essentially artificial -- when two or more players tie for first, none stood out so much to deserve victory. If you don't like tiebreaks - prove that you're the best, and win outright. Rapid or blitz tiebreaks, while letting the players decide, are at a different time control than the event itself. The system used in the candidates meant that the result was based on normal time control games only. A substantial benefit of the candidates tiebreak rules is that it sets up situations like we just saw in the 14th round - where both Carlsen and Kramnik had to play for a win - even though they were leading the event. If a rapid playoff was on the cards - I expect there would have been a lot less risk-taking / unexciting draws instead.
          Would you be in favor of the 3-1-0 point system being used, since this would make it more explicit that wins are valued more than draws?

          With the 3-1-0 system the final standings would have been (assuming no games ended differently because of the different scoring system):

          Code:
          Name				Fed		Elo    Pts
          --------------------------------------------------------------------------
          Magnus Carlsen			NOR		2872	22
          Vladimir Kramnik		RUS		2810	21
          Levon Aronian			ARM		2809	21
          Peter Svidler			RUS		2747	20
          Boris Gelfand			ISR		2740	15
          Vassily Ivanchuk		UKR		2757	15
          Alexander Grischuk		RUS		2764	14
          Teimour Radjabov		AZE		2793	 9
          Note how Ivanchuk moves past Grischuk even though Grischuk had 1/2 more point than Ivanchuk in the regular scoring system.

          But the main thing is that Carlsen would have won clear, and there'd be no controversy about tiebreaks. Of course, there WOULD be controversy about the 3-1-0 system being used.
          Only the rushing is heard...
          Onward flies the bird.

          Comment

          Working...
          X