How fast are the chess engines?

White locks the board up, at the start not one of his pieces goes into to black half of the board they stay behind the pawns. What is telling is when the engine sacs......move 144.....If you gave a human this position the human would see that black would need to sacrifice to break through whites pawn chain, the human would also see he could happily arange his pieces for this sac with no interference from the other side, basicly the human would pick a focal point of squares sac on it and have all of his other pieces ready in that general area.......the computer on the other hand cant plan to that level so when it sees the 50 move draw rule coming up it goes ......Normal move = 0.00......whats the next best move...sacrifice a rook =-2.34 (or something) so the best move for the computer is to blast the rook into a pawn rather pointlessly, just goes to show how dent chess engines can be at times.



On this note theres that famous Kasparov vs Fritz3D game where the engine cant find a plan and pushes wood around pointless until Kasparov cracks open the queenside, quite amusing seeing the 2800 odd computer fail to find the f5 pawn push breaking move that almost any 1500 would find.

Edit. Was trying to work out a sac for the position around move 100 (after the dark square bishop had gone) that would let black win but even after sacing both rooks and the queen then trying to promote a pawn doesnt seem to work.

But even as far back as move 40 when the dark square bishop is still on something simple like g5 would blast the position open and allow black and easy win.

Originally posted by Sicilian Smaug
If you did feed the moves of an in progress game into Fritz then [b] you would be the engine abuser.[/b]

I didn't say it clearly (takes me a while to get unambiguous). I was suggesting I put MY moves into Fritz and see if Fritz makes the same moves as my opponent. How can that be a violation?

When I started this thread by saying "I think I might be playing someone who..." I think it would have been more accurate to say, "I think, someday, I might encounter someone who". I am in several games right now and I do not suspect anyone of playing with an engine -- at least not one of Fritz's caliber (I tease).

Thanks to all for the feedback.

1... Rxc3 2. Bxc3 d2 3. Bxd2 Rc2 4. Kh3 Be2

d2 followed up by Rc2 seems to do the trick. The black queen is going to come into the game even if black has to play h4, Qh8, Qh5 etc.

If the white knight takes the rook then e5 sac a pawn or two and black should be able to enter on the light squares.

Originally posted by wormwood
who? man, he must've been one colossal monument of idiocy... - the lowest I've seen was 1800, and I think even that was pretty incompetent use of a perfectly good engine. it gives me physical pain to even think how bad a 1400 engine user must suck at chess.

Took me a minute to find him again on the removed players list.

I dont know how to put in the user number link deal, but the name was: dominic 0987

I guess he was pretty bad before using the engine, so getting over 1400 was a peak for him.

lol to the "colossal monument of idiocy" found that to be quite funny.

also, one of the recent players to be banned was just below 1800, but I guess he was well on his way to 2000 just got caught too soon.

So someone has to be targeted before they are looked at? I wondered about the cheating deal, why some ppl manage to cheat their ways to 2000+ over 600 games or so while others only make it 50 before they are busted.?

From chess today.

[Event "X3D Match"]
[Site "New York USA"]
[Date "2003.11.16"]
[Round "3"]
[White "Kasparov, G."]
[Black "X3D FRITZ"]
[Result "1-0"]
[ECO "D45"]
[WhiteElo "2830"]
[Annotator "Nikolai Vlassov"]
[PlyCount "89"]
[EventDate "2003.11.11"]
[Source "Chess Today"]
[SourceDate "2003.11.18"]

{After the previous game I suggested providing the Man with the low-power
software for such matches to insure him against blunders. Now it is time to
bring up the mirror proposal, and allow the Machine to consult a weak player
who will direct it from time to time. A second-grade player would suit fine to
turn the clown show into an intellectual encounter, in my opinion. However,
tastes differ.}
1. Nf3 Nf6 2. c4 e6 3. Nc3 d5 4. d4 c6 5. e3 a6 6. c5 Nbd7
7. b4 a5 $2 {Diagram # This is a very strange and weak move. White yearns for
b4-b5 but the a6 pawn prevents it, so Black is helping White! My Fritz 8 puts
this move into 5th place - not every software improvement leads to success in
the game!} ({After} 7... b6) ({or} 7... Qc7 {
Black's position has defence potential.})

8. b5 e5 9. Qa4 Qc7 10. Ba3 e4 11.
Nd2 Be7 {#} 12. b6 $1 $146 ({The following game saw} 12. Be2 h5 13. b6 Qd8 14.
h3 Nf8 15. O-O-O Ne6 16. Ndxe4 Nxe4 17. Nxe4 h4 18. Nd2 O-O 19. Rhg1 Re8 20.
Bd3 Bf8 21. Bb2 Ng5 22. Qc2 a4 23. a3 Qe7 24. Rde1 Ne4 25. Nf1 Qg5 26. f3 Nf6 {
Reshevsky-Keres, 1948 and White had a winning position.}) 12... Qd8 13. h3 O-O
({Better is} 13... Nf8 14. Nb3 Ng6 {- if White captures the a5 pawn, it will
be difficult for him to get rid of pins:} 15. Nxa5 O-O 16. Qb4 (16. Bb4 Nd7)
16... Nd7) 14. Nb3 Bd6 $5 {
# This is the only interesting move made by Fritz in this game!} 15. Rb1 ({
After} 15. cxd6 $4 Nxb6 {the white queen is trapped.})
15... Be7 $4 {Admirers of the scientific and technical progress were disappointed - Fritz demonstrated
all its idiocy! It simply does not understand what his pieces are meant for
and what to do with them.} ({Bad is} 15... Nh5 $2 {in view of} 16. Nxd5 $1 cxd5
(16... Ng3 17. cxd6 Nxh1 18. Ne7+ Kh8 19. Nc5 $18) 17. cxd6 Nxb6 18. Qb5 Bd7
19. Qc5 Na4 20. Qc7 $16 {and White has an upper hand that's why it is
necessary to deprive White of the c7-square:}) (15... Ne8 {
now the following does not work} 16. Nxd5 $2 cxd5 17. cxd6 Nxb6 18. Qb5 Bd7 19.
Qc5 Na4 $17 {Black seizes the initiative.}) ({
My Fritz prefers quite a human move} 15... Bb8 {- the bishop is transferred to
the diagonal where it supports the ...f7-f5-f4 advance, which gives Black
counter-play.}) 16. Nxa5 Nb8 {Steinitz was right - the initial position is the
best for the pieces and the Machine starts to realize it! :-) Computer only
needs to return the bishop to f8 as it did against Kramnik!} 17. Bb4 Qd7 18.
Rb2 $1 {Diagram # An excellent prophylactic move! Kasparov defends the f2 pawn
and prepares to transfer his king to the safe queenside.} Qe6 (18... Qf5 {
is impossible due to} 19. Nxc6 $1 $18) 19. Qd1 Nfd7 20. a3 Qh6 21. Nb3 {
Having captured the a5 pawn, White's pieces stepped back and cleared the way
for his a-pawn. This plan is obvious for a man, but is beyond machine's
comprehension because it is too long-term.} Bh4 {
Threatening to take on e3. In the previous game a similar trick worked!} 22.
Qd2 {#} Nf6 $2 ({Fritz does not understand (by the way, Junior understands)
that it is necessary to create counter-play with} 22... f5 {No machine is able
to think like human and to realize where pieces should be. For instance, if we
put the black rook on f7, then transfer the knight along f8-e6-d8 (if
necessary) and make it harder for White to breakthrough on the queenside,
White will be unable to attack on the kingside because his pieces are weaker
there. It would be a draw!}) 23. Kd1 Be6 ({The following does not work} 23...
Bxf2 $6 24. Qxf2 Ng4 25. Qd2 Nxe3+ 26. Ke2 Nxf1 27. Rxf1 $18) 24. Kc1 Rd8 {
Here is the overprotection of the d5 point - Nimzowitsch would have been
happy! Now the knight should be moved from b8 to e7. You will laugh, but this
is exactly what Black is goingtodo!} 25. Rc2 Nbd7 26. Kb2 Nf8 27. a4 Ng6 28. a5
Ne7 {# If Fritz had not made his last six "developing" moves, his position
would have been much better.} 29. a6 bxa6 30. Na5 Rdb8 31. g3 $5 {This delay
s Garry's desire to involve his h1 rook in the game. However, he has already
had enough forces.} ({Winning is} 31. Na2 {
and Black is unable to protect his pawns on a6 and c6.} Bc8 (31... Ne8 32. Ka1
(32. Bc3 Nc7)) (31... Nd7 32. Ba3 Bf6 33. Nb4 Nxb6 34. cxb6 Rxb6 35. Ka2 c5 36.
Rxc5) (31... Bg5 32. Ba3 Nd7 33. Ka1) 32. g3 Bg5 33. h4 Ne8 34. Bc3 Nc7 35. Nb4
Nb5 36. Bg2 Bf6 37. Naxc6 Nxc6 38. Nxc6 Rb7 39. f3 {However, Kasparov's plan
is better because it does not involve any complications.}) 31... Bg5 32. Bg2 ({
Not so convincing is} 32. Na2 Bg4 33. Rg1 Bf3 34. h4 Ng4) 32... Qg6 33. Ka1 Kh8
$4 (33... Bc8 {33...Bc8 was necessary, followed by moving the knight from f6
to d8. But Fritz does not realize this.}) 34. Na2 Bd7 35. Bc3 Ne8 36. Nb4 Kg8
37. Rb1 Bc8 38. Ra2 Bh6 39. Bf1 Qe6 40. Qd1 Nf6 41. Qa4 Bb7 {
The time control has been reached and now it's time for action:} 42. Nxb7 Rxb7
43. Nxa6 Qd7 44. Qc2 Kh8 45. Rb3 {# For some reason Black resigned here. Of
course, I understand that White is winning but the board is full of pieces,
only one pawn is missing, a man can blunder when the forces clash. It is at
least strange to break the show here.} 1-0

With your Red Hot Pawn rateing I am guessing that Fritz could beat you with a min. set for the game time.

Originally posted by Shinidoki
....and the rules of RHP state we are allowed to use opening books.

Surely this is cheating if you use a book during the game?
By the way, where are the RHP rules located (I'm new in town)?

Originally posted by Mixo
Surely this is cheating if you use a book during the game?
By the way, where are the RHP rules located (I'm new in town)?

(b) You will not use chess engines, chess software, chess computers or consult any third party to assist you in any game. Chess books and databases can be consulted during play

http://www.chessatwork.com/myhome/termsofservice.php

Originally posted by Turanthor
Took me a minute to find him again on the removed players list.

I dont know how to put in the user number link deal, but the name was: dominic 0987

I guess he was pretty bad before using the engine, so getting over 1400 was a peak for him.

lol to the "colossal monument of idiocy" found that to be quite funny.

also, one of the recent players ...[text shortened]... their ways to 2000+ over 600 games or so while others only make it 50 before they are busted.?

type [*uid]72095[/*uid] without the asterisks, and it'll look like:

User 72095

oh dear. that guy really had no idea what he was doing. not even after 800+ games played and using an engine! oh boy... I'm just stunned.

yea, people have to be reported to the game mods before they are investigated. there just isn't resources to check all players, or even a significant number of us properly.

the 'old che@ters' probably got 'established' as strong players before there were game mods, and not so many players are able to spot them just by playing them. but now, everybody knows what's what, and a sudden rise of 200+ points probably triggers a mass of complaints, resulting in investigation.

just like happened to 'wuffell' User 195025. -he played in the same tournament as I, the april banded 1600-1699, and I checked his graph when it started. the guy had been 1600'ish for hundreds of games, and suddenly jumped to 1800. he also went from a couple of dozens of games to 100+ at the same time. I thought it was suspicious, but maybe that he was dragging all his lost games or something. then he started to rise more, and I started waiting for his name to be overlined. it took a while, but there it is now. 3b.

people can also rise (at least) 200 points suddenly without che@ting, but there usually are telltale signs when you take a close look at their stats or play with them. wins/losses ratio, how long they've been on rhp, what kind of 'scalps' they've gathered (excluding timeouts & strong players dropping pieces). sometimes a graph can even look suspicious, but then you notice the guy plays seldom, and the games stretch over a scope of years. then there are the likes of jusuh & bowmann, who are obviously strong players but just don't play enough games here to get their rating anywhere near where it would end up.

Originally posted by techsouth
I wonder how hard it would be to measure the length of time to make each move as a reason to suspect engine use.

For example, any human, even playing well would find themselves in some positions for which the move is pretty obvious, and others that could take longer. You would expect a human to have high variance in move times. Whereas using an engine ...[text shortened]... es, but I fully expect that some engine users would use engines for some moves, but not others.

IN fact I am a quite speedy mover (occasionally comitting large overviews 🙂 Would you ban me on engine use?

Originally posted by Sicilian Smaug
The idea of linking Engine use with times between moves in correspondance chess is totally bizarre.

yes it was an incredibly stooopit suggestion

Originally posted by Ponderable
IN fact I am a quite speedy mover (occasionally comitting large overviews 🙂 Would you ban me on engine use?

It would be absurd to use move times alone as an indicator of engine use. What I suggested was to analyze move patterns as an automatic way to pick people who need more careful attention.

Edit:
I also noted that your rating is around 1200. If you sent an e-mail confessing to engine use, I'd probably not believe it. Being a quick mover, even for each and every move, and being rated 1200 is not suspicious at all.

Edit #2:
Also, I was never suggesting that "faster moving = more suspicious". It is more the consistency of moving time in spite of large variance in board complexity. Personally, I would need a minute each move just to get the position entered into the computer, and another couple of minutes of engine running time to make computer use worthwhile. I am sure many many players consistently move faster than would be likely for an engine user. Also, even engine users have to do other things. So time for a move is not really meaningful. But the time between seeing your opponents move and making your move may show patterns for engine users.

Originally posted by JokerFive
yes it was an incredibly stooopit suggestion

Is it really necessary to call my idea stupid? Do you make many friends that way? Have you fully understood my idea to the point you could discount it? If so, congratulations for being that much smarter than me such that you can discount it without even examining the data. But if this idea makes me stupid, then I guess that doesn't necessarily mean you're very smart, just by being smarter than such a stupid person.

If a person regularly views his opponents move and then always waits 2 to 3 minutes before making a reply (never making any quick moves, nor never taking much longer than 2 to 3 minutes), and continues this over a long period, personally I would at least begin to be suspicious. Also, if a person who played very well (I'm not talking about poor players who are bad because they're impulsive) never went to another game prior to moving on the game he just brought up, I would also be suspicious.

But that is just an idea. The moderators have dozens of banned players that have already been caught. If they kept data on when an opponent's move was first viewed and when their actual move was made, they may find patterns that are common amoung engine users that is rare amoung others.

I am not suggesting that a person be banned merely because their move times matched the move times of engine users. With more than 10,000 regular users, it would be nice to have an automated way to flag suspicious behaviour. None of these people would be banned without running whatever tests they already run as a result of tips.