mAGNUS CARLSEN VS houdini 2 PRO

Sort:
Rasparovov
FEDTEL wrote:
1MIN-Guardian wrote:

Carlsen would destroy Houdini if you give him 20 hours versus Houdini having 30 minutes.

Don't think so, Even if you gave houdini 1 min/game it will win more games than him.

I don't think that's correct. My Houdini always takes a minute or two to settle for a move. 

Rasparovov
FEDTEL wrote:
Rasparovov wrote:
FEDTEL wrote:
1MIN-Guardian wrote:

Carlsen would destroy Houdini if you give him 20 hours versus Houdini having 30 minutes.

Don't think so, Even if you gave houdini 1 min/game it will win more games than him.

I don't think that's correct. My Houdini always takes a minute or two to settle for a move. 

ofc, I know that the more time you give, the stronger Houdini will be, but do you know the ELO points decrease if you give houdini 1 min\game instead of 30 mins\game? I think it will be no more than 400 ELO.

the Idea is that the engine's strength (or number of moves searched) increases LOGARITHMLY with time (e.g. let Houdini evaluate a position for 1 sec, then for 60 secs, you will see the difference in depth reached is relatively very small between the two cases despite the second time is 60 times longer than the first).

now, can you Imagine the amount of decrease in strength for a Human playing at 30 mins\game compared to the same Human playing at 1min\game!

yes a computer will play a 1 against 1 minute game better but now it's 20 hours versus 1 minute. 

Tartarus_BW

Maybe he can win with chess 960?

Rasparovov
FEDTEL wrote:
Rasparovov wrote:

yes a computer will play a 1 against 1 minute game better but now it's 20 hours versus 1 minute. 

maybe, but 20 hours is too much (I meant 30 mins for Carlsen), even with 20 hours Magnus Carlsen can win some games ofc but not "destroy" houdini playing 1mins\game as mentioned above, let alone 30 mins for Houdini.

I'm pretty sure Carlsen would destroy Houdini 2 with 20 hours against 1 minute.

KeyserSzoze

end of this thread?

 

"The food arrives, and dishes crowd the little table. How does he feel about chess computers, I ask as we fill up our plates. “I can’t beat the best computers. They have complete information, so how could we expect anything else? I don’t look at computers as opponents. For me it is much more interesting to beat humans. Again, not a trace of angst.December 7, 2012

 

rest of the story here

mvtjc
Tartarus_BW wrote:

Maybe he can win with chess 960?

AGREE! Same case that until now noone has ever created an AI for Arimaa that can beat top players.Smile

Weakest00

btw houdini created by man not machine.....nothing is impossible......;)

Mandy711

Carlsen have a  to win if Houdini is handicapped. No opening book. no endgame tablebase, hardware using Pentium I processor with 64 MB RAM PC-133

pfren
Mandy711 wrote:

Carlsen have a  to win if Houdini is handicapped. No opening book. no endgame tablebase, hardware using Pentium I processor with 64 MB RAM PC-133

Any idiot would agree with you, hands down.

Spaskasjasky

Hello all.  The unique film about Magnus Carlsen, the worlds best player is available for viewing here:

www.princeofchess.com

ThaGoodBoy

but we have to accept the fact that all GM now use engines to analyse their finished games to know which moves was wrong,

in all cases it's normal that humans lose against engines, because

chess= calculations

calculations=calculator

calculator= engine

our brain is not made to calculate like those engines

Mandy711
pfren wrote:
Mandy711 wrote:

Carlsen have a  to win if Houdini is handicapped. No opening book. no endgame tablebase, hardware using Pentium I processor with 64 MB RAM PC-133

Any idiot would agree with you, hands down.

How about the experts? Anyone disagree? Correction on original post. Carlsen have a chance to win...

Clavius

You can cripple any hardware/software enough to make it weaker than a human player.  This is about as meaningful as disabling a drag racer car so that an Olympic sprinter can beat it in the quarter-mile.  Better just to think of humans and chess-playing computers as two different forms of the sport.  

I've always thought of the best chess games, no matter who or what played them, as a window on genius.  That is, a chance to see something of the process underlying the best "thinking" in a way not usually possible when looking at a scientific discovery or artistic creation.  It matters little to me whether the genius was a human player or software created by a human.

TheGreatOogieBoogie

I always hate it whenever people have all caps, but out of habbit comically lowercase someone's name because of the shift+ first letter in name instinct.  It makes words look too ugly.

scottsdetailcity

carl may have a Photographic Memory but a 'Model' ? That's just fruity !

Rasparovov
FEDTEL wrote:
Rasparovov wrote:
FEDTEL wrote:
Rasparovov wrote:

yes a computer will play a 1 against 1 minute game better but now it's 20 hours versus 1 minute. 

maybe, but 20 hours is too much (I meant 30 mins for Carlsen), even with 20 hours Magnus Carlsen can win some games ofc but not "destroy" houdini playing 1mins\game as mentioned above, let alone 30 mins for Houdini.

I'm pretty sure Carlsen would destroy Houdini 2 with 20 hours against 1 minute.

How can we know who's right?

( destroying a player is not defeating him\it in one match, it's defeating him\it in almost all the matches played)

Because there's statistics for how long Houdini can calculate to make a game, ~50 moves, in one minute and that is not enough to beat the best chess player in the world Magnus Carlsen with 20 hours, making almost no mistake at all.

EmperorPenguin

Most of this thread is speculation.  I would like Magnus to post a game he had with Rybka or Houdini so we can all see the results.  Even better would be a live game for all of us to watch.  Although it would be nice to watch him either lose or even struggle over the board, it would be astounding if he could pull off a win.

Clavius

If you are playing against an engine in a correpondence game and not using an engine yourself, you will almost certainly lose.  If you are playing against a top engine-using correspondence player and only make the moves your engine recommends, you will also almost certainly lose.  

The best correspondence players know how to use their engines as a tool to explore alternative moves, using their own judgment in choosing the most strategically valid lines in complex positions.

ElKitch

I would like to see a match in which all these players team up:

 1  Carlsen, Magnus  g  NOR  2861  8  1990-11-30
 2  Kramnik, Vladimir  g  RUS  2810  8  1975-06-25
 3  Aronian, Levon  g  ARM  2802  8  1982-10-06
 4  Radjabov, Teimour  g  AZE  2793  0  1987-03-12
 5  Caruana, Fabiano  g  ITA  2781  11  1992-07-30
 6  Karjakin, Sergey  g  RUS  2780  11  1990-01-12
 7  Anand, Viswanathan  g  IND  2772  8  1969-12-11
 8  Topalov, Veselin  g  BUL  2771  0  1975-03-15
 9  Nakamura, Hikaru  g  USA  2769  8  1987-12-09

-----------against Houdini 3 on very good hardware.

Timecontrols: 3 days/move for the humans and 1 day/move for Houdini 3. The human players may deliberate together about the best move strategy, but not use an engine. Would Houdini still win?

pfren
jamesmichael wrote:

.I have Houdini 2 Pro , Strelka , Critter, StoackFish,RobboLito,Bouquet,Deep Saros, Deep Rybka  and Vitruvius,I think all together i have at least 188 Chess Programs.

...and you are still rated somewhere around 1000. Please consider removing a dozen of them and- who knows, you may reach 1100.