Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: GM John v/d Wiel vs Rebel prediction statistic

Author: Howard Exner

Date: 12:04:54 12/02/00

Go up one level in this thread


On December 02, 2000 at 14:27:16, Uri Blass wrote:

>On December 02, 2000 at 11:31:09, Ed Schröder wrote:
>
>>On December 02, 2000 at 09:02:00, Chessfun wrote:
>>
>>>On December 02, 2000 at 08:49:29, Ed Schröder wrote:
>>>
>>>>The correct URL is:
>>>>
>>>>http://www.rebel.nl/rebwiel.htm
>>>>
>>>>
>>>>On December 02, 2000 at 08:47:41, Ed Schröder wrote:
>>>>
>>>>>So far received 90 votes for the match prediction GM v/d Wiel vs Rebel:
>>>>>
>>>>>John v/d Wiel versus Rebel  0 - 6       2      2.2%
>>>>>John v/d Wiel versus Rebel  ½ - 5½      0      0.0%
>>>>>John v/d Wiel versus Rebel  1 - 5       0      0.0%
>>>>>John v/d Wiel versus Rebel 1½ - 4½      5      5.6%
>>>>>John v/d Wiel versus Rebel  2 - 4      16     17.8%
>>>>>John v/d Wiel versus Rebel 2½ - 3½     12     13.3%
>>>>>John v/d Wiel versus Rebel  3 - 3      18     20.0%
>>>>>John v/d Wiel versus Rebel 3½ - 2½     20     22.2%
>>>>>John v/d Wiel versus Rebel  4 - 2       9     10.0%
>>>>>John v/d Wiel versus Rebel 4½ - 1½      5      5.6%
>>>>>John v/d Wiel versus Rebel  5 - 1       0      0.0%
>>>>>John v/d Wiel versus Rebel 5½ - ½       1      1.1%
>>>>>John v/d Wiel versus Rebel  6 - 0       2      2.2%
>>>>>
>>>>>Number of predictions so far :  90
>>>>>
>>>>>GM John v/d Wiel will win    :  41.1%
>>>>>The match will end in a draw :  20.0%
>>>>>Rebel will win the match     :  38.9%
>>>>>
>>>>>Please subscribe your opinion to:
>>>>>
>>>>>http://rebel.nl/rebwiel.htm
>>>>>
>>>>>Added are a few small prizes for those who predict correct.
>>>>>
>>>>>Kind regards,
>>>>>
>>>>>Ed Schroder
>>>
>>>
>>>Can I ask what kind of payment incentive does VD Wiel get
>>>for winning 1) The match?, 2) Each game?
>>>
>>>I just want to find out what type of incentive exists for
>>>him to try to win say game 6 if he is up 3-1/2 - 1-1/2.
>>>
>>>Sarah.
>>
>>I am not sure if this free information. I will have to ask the
>>organizer first.
>>
>>Ed
>
>I think that this is important information for the prediction.
>
>I expect John v/d Wiel to prepare better if there is more money to earn from
>scoring more points.
>
>I hope that the payment is based on the principle that the money is proportional
>to the number of points that he get(He may get also money for playing).
>
>If he gets also money from winning the match than he has a reason not to try to
>do his best.

But many chess players, even at much lower levels than GMs, utterly despise
losing. They often times feel it is their right to dominate certain opponents. A
draw in many cases is viewed as failure. So motivation to win even with no
payment is high, especially from someone who has never lost to a computer in a
public event.
>
>He may try to draw if he leads 3-1 or 3-2 even if he has better chances.

That would still tell us something about the computer strength or weakness. That
is, if van der Wiel starts off with 3-1 then easily draws the next two games. I
believe a straight payment for a match win would be sufficient, just like any
other human vs human match.

>He also may try to win without objective basis if the result is 2.5-2.5 assuming
>that the difference in money between 3-3 and winning 3.5-2.5 is not the same as
>the difference between 3-3 and losing 3.5-2.5.
>I do not want it to happen.
>
>I hope that the payement is something like this:
>
>1000$ for 0 points
>2000$ for 0.5 points
>3000$ for 1 points
>4000$ for 1.5 points
>5000$ for 2 points
>6000$ for 2.5 points
>7000$ for 3 points
>8000$ for 3.5 points
>9000$ for 4 points
>10000$ for 4.5 points
>11000$ for 5 points
>12000$ for 5.5 points
>13000$ for 6 points.

This is the kind of match that meets all the requirements of a tough, fair
competition. Many have wanted the following in a computer-human match...

1. A tough, computer savy GM.
2. Standard time controls.
3. A series of more than one or two games to test how the human can adjust to
the computer opponent.
4. Having a human opponent that has access to the computer program beforehand,
unlike the Kasparov - Deep Blue contest where DB was shrouded in secrecy.
5. Decent playing conditions for the human player.




This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.