lichess.org
Donate

Human vs Computer

I know some guy who want to be defeated, but don't know anyone that can defeat him. Including chess engines. He want to stay anonymous and I will not reveal his identity. Apparently he can defeat any chess engine. He don't care about time, he says all players must have all the time they want. He says I can put the computer all day to work on 1 move, but I will not do that. I will not toast my computer to do that. He says someone who play fast don't like chess. Because if you play fast you make bad moves, and if you like to play bad moves you don't like the game. I don't agree with him, but whatever. He don't know openings, don't read chess books, don't use computer assistance, don't use annotations, etc. And don't cheat. I know him.

Here is an example, it is him, only his mind, no computer, no books, no annotations, nothing, 1 day per move. Against Komodo 5 64 bits, 512 MB of hash DDR3, Pentium Dual Core G620, Komodo alone, no opening book or endgame databases, 60 seconds per move.
About move 35 to the end, the guy played very fast, about 5 to 15 seconds per move. But I keep Komodo 60 seconds per move:

[Date "05/03/2015 19:27:02"]
[Round "1"]
[White "NN"]
[Black "Komodo 5"]
[Opening "Bishop's opening: Greco gambit"]
[Eco "C24"]
[Result "1-0"]

1.e4 e5 2.Bc4 Nf6 3.f4 d5 4.exd5 Nxd5 5.Ne2 Bc5 6.Nbc3 Nb6 7.Bb3 Qh4+ 8.g3 Qh3
9.Rf1 Qg2 10.f5 O-O 11.d3 a5 12.a3 N6d7 13.Ne4 Ba7 14.N2c3 Qxh2 15.Qf3 Nc6
16.Rh1 Nd4 17.Qf1 Nxc2+ 18.Bxc2 Qxc2 19.Qh3 Bf2+ 20.Nxf2 Nf6 21.Nce4 Rd8
22.Nxf6+ Kf8 23.Nxh7+ Ke8 24.Ng5 b6 25.Qh8+ Kd7 26.Qxg7 Kc6 27.Nxf7 Rd5 28.Nxe5+
Kb7 29.Rh8 Ka6 30.Nfg4 Rb5 31.Ne3 Qc5 32.Nd7 Qc6 33.f6 Bxd7 34.f7 Rg5 35.Rxa8+
Qxa8 36.Qxg5 Be6 37.Qe7 Bxf7 38.Qxf7 Qh1+ 39.Kf2 Qc6 40.Qc4+ Kb7 41.Qxc6+ Kxc6
42.Bd2 Kd6 43.g4 b5 44.Bxa5 Ke7 45.Bxc7 Kd7 46.Ba5 Ke6 47.g5 b4 48.axb4 Ke5
49.g6 Kf6 50.Rg1 Ke6 51.b5 Kd6 52.b6 Kc5 53.b4+ Kb5 54.g7 Kc6 55.g8=N Kb5 56.b7
Ka4 57.b8=N Kb3 58.d4 Kb2 59.b5 Kb3 60.b6 Ka4 61.Nc6 Kb3 62.b7 Ka3 63.b8=N Kb3
64.d5 Ka4 65.d6 Kb3 66.d7 Ka3 67.d8=N Kb3 68.Nd4+ Kb2 69.Nbc6 Ka2 70.Bc3 Ka3
71.Ra1#

As you see (you don't know him personaly as I do, maybe you don't belive me, but whatever) a human can defeat a top chess engine. Time is not a factor for him.
He played like Pele in football. Went to a big fight with tactical complications. Nice to know that people are still able to fight the iron monster.
I don't know why people don't try to defeat top chess engines anymore. You don't need piece handicap. People think they are invincible, but they are far from that.
Color me skeptical.

There's been no shortage of people who have played games and posted games beating engines over the last few years, but always with one common factor: no confirmation of the conditions.

They say they're not using an engine, but we have to take their word for it.

Here are the things we do know for sure:

1) The last highly publicized matches between humans and engines without odds ended with drawn matches as the best result for humans.

2) These matches were against engines that were much weaker (at least a couple hundred points) than today's engines. This is easily confirmed with the large volume of engine-engine testing done these days on rating lists.

3) The matches against current top engines with verified conditions have been at odds, and even with large odds, strong humans have struggled to draw.

So, with that known, we're supposed to believe that some random individuals are able to do what the world's best players have been unable to do for years.

The probability of that being true is akin to the probability that some random guy we've never heard of could consistently beat Lebron James in 1 on 1 basketball, or play better soccer than Messi and Ronaldo.

As with those hypothetical claims, I'll believe the claim when it's proven, but until its proven by being repeated with the conditions verified, I'm going to politely call shenanigans :)
To add to that, I should say that by no means am I saying current engines play perfectly.

They certainly do not, as evidenced by their losing games to one another still.

However, any claim of a human being able to beat a top engine is highly suspicious these days.

Also, in this particular case the game cited is very suspicious, as black, supposedly Komodo 5, made several huge tactical mistakes very early in the game.

Komodo 5 is pretty old, so I don't think I have a copy to test with, but I find it difficult to believe that it would play such moves.

I'll have to see if I can dredge up a copy of K5 to see if I can duplicate those move choices.

If I can't, then the most likely explanations are that 1) the game was manufactured or 2) Komodo was configured to play at less than full strength.
@OneOfTheQ I like skepticism. "There's been no shortage of people who have played games and posted games beating engines over the last few years". I'm no chess specialist, and I don't know that. Show me the games, I want to see them.
I only see people saying Houdini is invincible, bla, bla, bla.

You can download Komodo 5 for free here:

komodochess.com/downloads.htm

Test it for yourself. 60 seconds per move analysis. 2 threads. 512 MB of hash. I like skepticism, but I will not prove. Because I respect him and he want to stay anonymous. But I believe other people can do that, and I say if you want to see that happen close one guy that is very good in correspondence chess in a room with cameras and maybe you will see that. If you want a fair fight, you can say the game was not fair. Because Komodo only "think" 60 seconds per move. I think these games of Grandmasters against top chess engines are in equal time, no?
Hey Hellrazer.
For your own good don't take anything this guy purposes to you...
Seriously.

This topic has been archived and can no longer be replied to.