Partnership game - hypothetical

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

  • Partnership game - hypothetical

    This question was posted on another forum I read:

    You have white and are playing against a 2000 player. Every other move the world champ moves for you. No communication though. Your opponent realizes this. You would be favored as long as your rating is above what?

    I honestly cannot decide what the answer would be.

  • #2
    Re: Partnership game - hypothetical

    There would be strategy involved here. The 2000 player would try to control the game so that the GM would be forced to make the obvious recaptures, and his partner would face more complicated situations. Likewise the "team" would be trying to reverse this. Do we assume that the world champ plays perfectly - i.e. always choosing the best move?

    My response would be: 2000.

    Comment


    • #3
      Re: Partnership game - hypothetical

      I once played in a blitz event alternating moves with a partner. Team strength is a LOT weaker than the stronger player and may even be weaker than that of the lower rated player. (so your rating would need to be some amount over 2000)

      One of those blitz games was against Hikaru Nakamura. It was fairly evenly matched (we won a piece early, kept it for awhile, then lost it and the game). Incredibly frustrating for both Nakamura and I. Of course, me versus Nakamura solo - it's clear who the favourite would be :-)
      Last edited by Roger Patterson; Wednesday, 21st April, 2010, 01:49 PM.

      Comment


      • #4
        Re: Partnership game - hypothetical

        I am going to make a few assumptions:

        1) That the two partnering players realize the strength of their partner.

        2) That the world champ doesn't know anything other than the strength of their partner (i.e. he hasn't discussed that player's opening repertoire, style, etc.).

        3) That the world champ is allowed to make his moves based on the partnership, i.e. he might choose an inferior move because it would allow his partner to find relatively simple moves.

        4) That it is a reasonably slow time control game. Say at least 30 mins each side.

        I would say 1700-1900. If someone played like the 1700 version of Topalov, for example, then partnering with the actual Topalov would probably produce a much-improved 2000+ version of Topalov.

        I once did a tandem simul with Deen Hergott in Saskatoon. We did alright (something like 85%), but it was much worse than either of us would have done individually. Mostly that was because in position X he would move a piece forward, then on X+1 I would put it back. ;-)
        "Tom is a well known racist, and like most of them he won't admit it, possibly even to himself." - Ed Seedhouse, October 4, 2020.

        Comment


        • #5
          Re: Partnership game - hypothetical

          Originally posted by Tom O'Donnell View Post
          3) That the world champ is allowed to make his moves based on the partnership, i.e. he might choose an inferior move because it would allow his partner to find relatively simple moves.
          I don't know of any chess tournament where a player (GM or otherwise) isn't allowed to make inferior moves.

          Originally posted by Tom O'Donnell View Post
          I would say 1700-1900. If someone played like the 1700 version of Topalov, for example, then partnering with the actual Topalov would probably produce a much-improved 2000+ version of Topalov.
          I don't think I could agree with this. Remember, no communication. The 1700 Topalov isn't going to be able to grasp the more complicated moves of the GM Topalov, and will take some extra time to try and figure them out. Besides the extra time, the 1700 Topalov will sometimes get it wrong and interfere with what the GM Topalov is doing, and I think this would lead to a loss of coordination that would eventually be disasterous.

          But this could be tested perhaps? Not with any human, but with an early version of Rybka partnered with the latest version, playing in this format. Play them several matches against another chess engine, then play the early version of Rybka alone against the same chess engine for several matches (everything else being equal) and compare rating performances.
          Only the rushing is heard...
          Onward flies the bird.

          Comment


          • #6
            Re: Partnership game - hypothetical

            I predict 1600

            Comment


            • #7
              Re: Partnership game - hypothetical

              Originally posted by Paul Bonham View Post
              I don't know of any chess tournament where a player (GM or otherwise) isn't allowed to make inferior moves.

              ...

              But this could be tested perhaps? Not with any human, but with an early version of Rybka partnered with the latest version, playing in this format. Play them several matches against another chess engine, then play the early version of Rybka alone against the same chess engine for several matches (everything else being equal) and compare rating performances.

              Uh, no. How could one program the stronger Rybka to make inferior moves based on the knowledge that the "other" Rybka may not be able to understand what's going on?
              "Tom is a well known racist, and like most of them he won't admit it, possibly even to himself." - Ed Seedhouse, October 4, 2020.

              Comment


              • #8
                Re: Partnership game - hypothetical

                Originally posted by David Ottosen View Post
                This question was posted on another forum I read:

                You have white and are playing against a 2000 player. Every other move the world champ moves for you. No communication though. Your opponent realizes this. You would be favored as long as your rating is above what?

                I honestly cannot decide what the answer would be.

                It depends. The players have to understand the idea behind what the other player is doing. Assuming the world champ is helping a class B or lower player the 2000 players probably has the advantage. Also, who moves first. The regular player or the WC? Going through the opening on the white side, the player who plays first gets move 1, 3, 5, 7, and 9 as white and the other only 4 of the first 9 moves.
                Gary Ruben
                CC - IA and SIM

                Comment


                • #9
                  Re: Partnership game - hypothetical

                  Originally posted by Tom O'Donnell View Post
                  Uh, no. How could one program the stronger Rybka to make inferior moves based on the knowledge that the "other" Rybka may not be able to understand what's going on?
                  Elementary, my dear Watson! You simply program the stronger Rybka to make inferior moves based on the knowledge that the "other" Rybka may not be able to understand what's going on! :D

                  Ok, sorry for the humour... so what you seem to be saying, then, is that if a GM were paired with a 1700, and the GM were making inferior moves so that the 1700 could understand what's going on, the two of them together would play stronger than the 1700 would all by himself or herself. Is this correct?

                  Then of course, this couldn't be tested with chess engines. What the chess engine test would do is test whether a weak engine partnering with a much stronger engine would play better than the weak engine alone, without the stronger engine making any inferior moves. I still think that would be an interesting test.

                  Your test would have to be done with humans, and I would think it would be hard on the GM to try and figure out what move can a 1700 figure out. And if you switched the GM's partner to a 1900, how would the GM then change his or her play... very hard on the GM, I would think.

                  Even with the GM dumbing down, I dont' think it would work out. Somewhere there's going to be a misunderstanding, I would guess. I like the theory that the two of them together would play weaker than either of them alone.

                  Now, where was I.... ah, yes, changing Rybka source code....

                  if( partnerRating == 1700 )
                  {
                  playInferiorMove();
                  }
                  Only the rushing is heard...
                  Onward flies the bird.

                  Comment


                  • #10
                    Re: Partnership game - hypothetical

                    Originally posted by Tom O'Donnell View Post
                    , but it was much worse than either of us would have done individually. . ;-)
                    i.e. team performance was lower than the strength of the weakest player (leaving aside which person that was --:-;)

                    Even dumbing down doesn't work. You imperfectly speak the same language but miscommunication remains (I thought you were talking about seals. What! you thought I was talking about sex? [ask a french speaking person for an explanation]

                    In that same tandem blit tournament the following happened to me [only the important pieces are on the diagram - I remember the theme but not the exact details We were up a piece but I'm not too sure where all of whites pieces were]



                    we were black. My plan for this position was to play Qa7 and if RxN then Bc5. But... my partner had the move and he saw something. He joyfully plays 1.....Nh3+ [white plays 2. Kh1].

                    My turn. Well what to do. Don't want to move the Q and leave the Nh3 hanging. So 2.... Nf2+ [white plays 3. Kg1].

                    My partner glares at me and Smash! 3...... Nh3+ (so much for no communication between partners). [4. Kh1]

                    Well what to do. Nf2+ is a triple rep. We were up a piece so I moved the Q somewhere (and dropping the Nh3) expecting the open white K to give us an edge.

                    Man, was my partner upset. Stupid master who doesn't even know the basic mates.
                    Last edited by Roger Patterson; Wednesday, 21st April, 2010, 08:49 PM.

                    Comment


                    • #11
                      Re: Partnership game - hypothetical

                      [deleted not to confuse]
                      Last edited by Egidijus Zeromskis; Thursday, 22nd April, 2010, 01:13 PM. Reason: [deleted not to confuse]

                      Comment


                      • #12
                        Re: Partnership game - hypothetical

                        Roger:

                        From your description, your partner appears to be rather ... . I would think that anyone in his situation would ask something like "Okay, my partner is X00 points stronger than I am, what am I missing?" But maybe that's just me.

                        Anyway, that is blitz. I was thinking more for a "real" tournament game (more time for nuanced reflection).

                        Let's say you, Roger, are 2200. Let's say you are going to play a game with a time control that averages 3 mins/move. How well do you think you would play at 1 min/move? 1900? So a possible strategy would be to play all of your moves at an average rate of 1 min/move thereby giving your partner an average of 5 mins/move to play his. Assuming he is an 1800 player who would use the extra time, I don't see how the result would be worse than the 1800 playing alone.

                        This thread does give me that idea of suggesting to the RA Club a Thursday night event where the two players cannot communicate, the games are pretty slow (say 90 mins/side) and the sum of the two players' ratings cannot be over X (say 4000). I would imagine that two 1950s would beat a 2450 and a 1450, but maybe not.
                        "Tom is a well known racist, and like most of them he won't admit it, possibly even to himself." - Ed Seedhouse, October 4, 2020.

                        Comment


                        • #13
                          Re: Partnership game - hypothetical

                          the point of the anecdote is that there IS a Nf3.

                          Comment


                          • #14
                            Re: Partnership game - hypothetical

                            Originally posted by Tom O'Donnell View Post
                            Roger:

                            I would imagine that two 1950s would beat a 2450 and a 1450, but maybe not.
                            well the tandem event I was at was at one of the Gilbralter GibTel events and a rule was that the average rating could not be higher than x . Not quite sure in memory what x was but either 2200 or 2300. So there were lots of teams like Nakamura + someone, Sasikiran + someone etc. They all got smoked. Badly. I think the winners were Marc Caselais and J. Dworakowski (sp?) both whose ratings were close the maximum average allowed. [they either came first or came second behind a team with a similar set of ratings]

                            Comment


                            • #15
                              Re: Partnership game - hypothetical

                              Originally posted by Tom O'Donnell View Post
                              This thread does give me that idea of suggesting to the RA Club a Thursday night event where the two players cannot communicate, the games are pretty slow (say 90 mins/side) and the sum of the two players' ratings cannot be over X (say 4000). I would imagine that two 1950s would beat a 2450 and a 1450, but maybe not.
                              Tom, if this event does take place, could you try and get all game scoresheets and provide pgn files of the games, including who is moving on even and odd moves in each game? I know I would be fascinated to see that, and probably others too. It's a very intriguing idea.
                              Only the rushing is heard...
                              Onward flies the bird.

                              Comment

                              Working...
                              X