Authors of chess engines getting crushed by their creations

[MikeB]

Creation is too strong a word. The creative person is the one who first did something. The later ones follow a recipe.

Bob Hyatt, the Crafty author, was a pioneer of parallel computing.

Myself when i started Pulsar didn't have much claim to fame i was not following a recipes. But I wanted to do something different so I made a game plex. Pulsar for iOS and Android plays Chess, Atomic, Crazyhouse, Three Checks, Loser's Chess and Giveaway. And i spent time with all of them working on how to teach the chess code to play variants. I did start with chess code some parts stripped out.

So i think there is different types of creation even if following an area others are working in.

I wish to make a distinction between going down an unproven path that nobody has done before, and doing something that has been done before. The former is uncertain and risky and coming out the other side with something functional is the creative achievement and the creative process.
Going down a path that has already been proven, whether or not the problem solving involved was redone or looked up, is an entirely different affair.
It takes a very special personality and determination to go where nobody else has been, and we need/use a special word to describe it. That word is creativity. It doesn’t apply to something that’s already been knowingly done elsewhere.

I think "pioneer" is the word you are looking for. Pioneers are the first ones to go down an untraveled trail. There is no question, Bob was one of the first early pioneers of computer chess.

[Henk]

Stupid tactics.

[pgn] [Event "Computer Chess Game"] [Site "HP"] [Date "2019.05.05"] [Round "-"] [White "SchaakMachine"] [Black "Henk"] [Result "1-0"] [TimeControl "420"] [Annotator "1. +0.56"] 1. d4 {+0.56/8} d6 2. e4 {+0.71/8 8} e5 3. c3 {+0.75/7 8} Be7 4. Nf3 {+0.88/7 8} Nd7 5. Bb5 {+0.66/8 8} c6 6. Bd3 {+0.78/8 8} Ngf6 7. O-O {+0.70/8 7} O-O 8. Nbd2 {+0.67/8 7} Re8 9. h3 {+0.72/8 7} Bf8 10. g4 {+0.56/8 7} g6 11. Qc2 {+0.65/8 7} Bg7 12. Nc4 {+0.84/7 7} Qc7 13. Bd2 {+0.80/7 7} b5 14. Na3 {+0.37/8 6} d5 15. dxe5 {+0.35/8 6} Nxe5 16. Nxe5 {+0.46/9 6} Qxe5 17. exd5 {+0.20/8 6} Nxd5 18. Rfe1 {+5.40/9 6} {Black resigns} 1-0 [/pgn]

[bob]

Wasn't intending to make this anything about me. However, for the record, my first chess program played its first move in very late 1968. I was a computer science major at the University of Southern Mississippi and star trek / spock / computer chess had already caught my interest. There was practically nothing one could find on the topic other than Shannon's original paper and then somewhere in the late 60's the Greenblatt paper in the DEC journa and then the COKO IV paper in the ACM journall. I think back then EVERYBODY was creating things left and right.. Today, the best "creating" seems to be the resurgence of ANN approaches, which still has some catching up to do. I can think of lots of significant "creations". Beal's null-move concept. Slate's (chess) and Samuels (checkers) bit board stuff. Forward pruning ideas like LMR, LMP, futility, etc. Extensions from the simple escaping check through singular extensions (Hsu). And even combining ideas is a form of creation.

None of which really has anything with the notion that probably every chess program that exists today can beat its author soundly, unless the program has lots of gross bugs.

BTW I have not "been gone" and am "coming back". Been enjoying retirement, traveling in our RV with my wife, running up and down the river fishing, etc. Biggest decision of any day is "do I shave today or not?"

[adams161]

I"ve always had a firm respect for the break throughs on the computer science level in chess since I first became familiar with the scene in late 1998 on ICC's channel 64 for chess programmers. Back then Bob was online running Crafty(C), Diepeveen was running Diep, Bruce Moreland was running Feret which was then the 1996 world championship program. And a ton of others in the cast like KiwiPete and many others.

But what i want to mention in terms of creation is that we are not just creating chess algorithms but there is a field also of creating the game. Pulsar started it's first new game, Three Checks, in the second half of 2002. Now there was no Three Checks program that I was aware of or web page to lookup this is how one does Three Checks. But I just thought about it and said how can I make a chess engine play three checks. Seemed obvious after awhile. Just score checks in evaluate. And i had to hash them into the board. But it was never obvious what the scores should be for first and second check and a lot of games went into deciding that After that, took on atomic, and it had it's own issues like this concept of a loose rook that can always be within one move of checkmate. And of course more chess program work to extend hash now to atomic captures. And again I only heard of one other atomic program at the time, OllieThink, and i never saw its code and not sure if i even ran it. There were not web pages for how to code atomic chess. And this continued as i moved on to loser's and Crazyhouse etc.

I see still a lot of game making going on in chess. Variants that don't use the 8x8 board have started becoming common. And when new games are made or new games have their first engines coded for them there are creative challenges that don't have a book or web page to look it up in. But to be fair this is probably just a fraction of the new creation going on today in Chess. We still have pioneers i think.

[MikeB]

Wasn't intending to make this anything about me. However, for the record, my first chess program played its first move in very late 1968. I was a computer science major at the University of Southern Mississippi and star trek / spock / computer chess had already caught my interest. There was practically nothing one could find on the topic other than Shannon's original paper and then somewhere in the late 60's the Greenblatt paper in the DEC journa and then the COKO IV paper in the ACM journall. I think back then EVERYBODY was creating things left and right.. Today, the best "creating" seems to be the resurgence of ANN approaches, which still has some catching up to do. I can think of lots of significant "creations". Beal's null-move concept. Slate's (chess) and Samuels (checkers) bit board stuff. Forward pruning ideas like LMR, LMP, futility, etc. Extensions from the simple escaping check through singular extensions (Hsu). And even combining ideas is a form of creation.

None of which really has anything with the notion that probably every chess program that exists today can beat its author soundly, unless the program has lots of gross bugs.

BTW I have not "been gone" and am "coming back". Been enjoying retirement, traveling in our RV with my wife, running up and down the river fishing, etc. Biggest decision of any day is "do I shave today or not?"

+1 Glad to hear!

[bob]

Crap. Meant to write "have not been gone, am not coming back." That is, I never left. just haven't had much to say. It's amazing how busy life gets when you own a boat, an RV, automobiles, house, have family, travel, most everything is "do it yourself" and so forth. Sometimes I am busier now than before I retired. I can guarantee you however, I would NEVER "go back" and un-retire. enjoying life way too much for that...

[Leo]

lol. Very good.

[JohnWoe]

I played a chess960 blitz game against Sapeli 1.03.
I haven't myself played lots of against my own engine.
I usually watch how it plays against other engines instead of numbers.

[pgn][Event "Computer Chess Game"] [Site "pc"] [Date "2019.05.11"] [Round "-"] [White "me"] [Black "Sapeli 1.03"] [Result "0-1"] [TimeControl "40/180"] [Variant "fischerandom"] [FEN "nbrnkqbr/pppppppp/8/8/8/8/PPPPPPPP/NBRNKQBR w HChc - 0 1"] [SetUp "1"] {-------------- n b r n k q b r p p p p p p p p . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P P P P P P P P N B R N K Q B R white to play --------------} 1. Nb3 Nb6 {+0,06/12 6} 2. Nc3 Nc6 {+0,03/12 6} 3. f3 O-O-O {+0,04/12 6} 4. e4 e5 {+0,01/11 5} 5. Be3 d6 {+0,00/11 5} 6. Qf2 Nc4 {-0,19/13 5} 7. Nd1 f6 {-0,27/11 5} 8. d3 Nxe3 {-0,28/12 5} 9. Nxe3 Be6 {-0,29/12 5} 10. O-O a6 {-0,34/10 4} 11. c4 Ba7 {-0,36/10 4} 12. Kh1 h5 {-0,42/10 4} 13. Qe2 h4 {-0,48/11 4} 14. Nd5 h3 {-0,54/11 4} 15. g3 Nd4 {-0,53/11 4} 16. Nxd4 Bxd4 {-0,53/10 4} 17. b4 Qe8 {-0,53/10 4} 18. Bc2 Qh5 {-0,49/9 4} 19. a4 c6 {-0,52/9 4} 20. Ne3 Qg5 {-0,47/10 4} 21. Rfe1 Bc3 {-1,51/11 4} 22. Rf1 Bxb4 {-1,50/10 4} 23. g4 Bc5 {-1,55/10 4} 24. Nf5 g6 {-1,63/11 4} 25. Ng3 Rh7 {-1,61/11 4} 26. a5 Kb8 {-1,62/9 4} 27. Ba4 Be3 {-1,62/9 4} 28. Rb1 Rc7 {-1,62/10 4} 29. Rfe1 Bd2 {-2,54/13 4} 30. Red1 Bxa5 {-2,53/12 4} 31. d4 exd4 {-2,64/10 4} 32. Rxd4 Qc5 {-2,64/10 4} 33. Qd3 Bb6 {-2,62/10 4} {White resigns} 0-1 [/pgn]