Is there any chance that a 1300 rated player can beat a 2700 rated player?

  • #5361

    A question that is asked this weekend; Can a 14 year old 2000 kid, Daniel Norquelle, beat the 2800 Fabiano Caruana in the Norway Gnome match? Or can a 2200 beat the world Champion?

  • #5362
    chessking1976 wrote:
    Mauve26 wrote:

    the answer is yes.

    Sometimes you only want to "like" a comment, not reply to it.

    Then you must like this comment, because you certainly did reply to it.

  • #5363

    The short and simple answer to this is:

    Yes - chess games are determined by many unknown and uncontrollable factors - brain chemistry can make you play good one day, and bad the other day. On the more extreme level, if the 2700 has a heart attack or stroke during the game, the 1300 can win (or something like that happens). 

    This is just an extension of "good form, bad form", which allows for many upsets in chess to happen. 

  • #5364

    Heart attack during play?

  • #5365
    Sometimes I take a knife into online games so I can stab the opponent when they're winning.
  • #5366

    Or, if you'd rather, some hallucination brought about by extremely bad form. Remember Kramnik blundered M1. Basically, the answer is yes, because it's how weak players beat strong players occasionally, and so if we extend those conditions, then it's possible for a player's form to be so bad they blunder and lose to a 1300 (and the 1300 could be very lucky, and be in good form). 

  • #5367

    possible, but not plausible.

     

  • #5368

    Definitely not plausible for a 2700 to beat a 1300, I agree. I would buy a lottery ticket rather than bet on that

  • #5369

    I think chances are probably 1 in 50,000. 

  • #5370

    is there a chance of raining in the desert?

  • #5371

    Well, according to the Elo system a 200 points lower player has 25% odds of winning. Here it is 1400 points. The math is simple: 0.257≈0.00006 or 0.006 %. However, I think this math won't work because when you have such a huge difference it will be further augmented by these 1400 points. So much so that the odds are less than one out millions (or something of that magnitude) under standard time control. This 0.006% (1 win out of 60.000) may only work in bullet, I guess, where a strong GM might still blunder. However, a low Elo player must be no slowpoke in bullet or he/she won't convert even a serious blunder against a GM. All this math and reasoning will collapse and be no longer valid if you pair a 1300 Elo player with a 2500 engine. A human 1300 will stand a chance of probably less than one out of a billion!!! And of course even less than that in a bullet game against an engine happy.png

    This math must also be no good because if they play so many games a 1300 player will learn a lot along the way, improving his/her Elo rating happy.png

  • #5372

    Funny thing is a 1300 Elo player stands no chance against a 2700 GM even if that GM has only one minute per game and a 1300 player has as much time as he/she wants! That's because the GM will also compute and evaluate positions while a 1300 player is thinking. Even if a 2700 player plays blindfolded, a 1300 stands no chance happy.png

     

  • #5373

    As far as the theory of probability goes, it's interesting to calculate a ballpark figure of what odds are that a 1300 player will beat a 2700 Elo engine in a bullet game? It's mind boggling.  Don't know how to evaluate that probability happy.png  Only meaningless guesses come to mind with meaningless numbers like10-1000 or 10-1000000.  Such nonsense happy.png

     

  • #5374

    Truth be told, a ballpark figure of the odds that a 1300 player will beat a 2700 Elo engine in a bullet game are not mind boggling. Only extremely approximate numbers are possible to obtain here but still possible and even easy happy.png

    Assume the probability of a good move that a 1300 makes (if all moves are like that) that can eventually win against a 2700 Elo engine is 0.05. Then assume you have to make 80 such moves to win against a 2700 Elo engine (the engine defenses ferociously extending the game to many moves e.g. Fritz 6 against Fritz 11. That's why we assume 80 moves). The probability (extremely approximate number) is 0.0580 = 8x10-105

    No big deal even here. We can still make rough estimates. The margin of error is huge here, the probability can be 10-90 or it can be 10-120. So the estimates are not very good. However the probability cannot  be as low as 10-1000 happy.png

     

  • #5375
    Mauve26 wrote:
    chessking1976 wrote:
    Mauve26 wrote:

    the answer is yes.

    Sometimes you only want to "like" a comment, not reply to it.

    Then you must like this comment, because you certainly did reply to it.

    True

Top
or Join
=

Online Now