Sunday, December 2, 2007

Oddities With Chess Publisher

If you have tried to follow my game against Anatoly Tsinker, you will see a very strange thing happen. For blacks' 18th move of Ng4, it takes the knight from a4 and moves it to g4. Then for black's 19th move of Bf6, it plays Bxf6. (It's capturing it's own knight.) Move 20 for black is supposed to be Kc3, forking my rooks. Move 21 is supposed to be Nxa2. Then the rest of the game seems to play normally. I checked the pgn file, and the moves are fine. It plays through normally on Chess Base. I even reloaded it on Chess Publisher and edited the post with the new link. Same thing! I'm too tired and it's past my bedtime so it's staying as is for the moment.

Anyone else seen this happen before?

12 comments:

Glenn Wilson said...

I have not used Chess Publisher but I wonder if using long algebraic for the move(s) it is messing up would fix the problem. So, change you pgn to have 18 ... Nfg4 or maybe 18 ... Nf6-g4.

I suspect the software consists of a "parser" of the pgn that generates the list of moves to make and the "parser" has a bug..."disambiguating" your 18th might work around the bug....

Polly said...

I've never encoutered the problem before. Though maybe I have, and just haven't noticed. But it seems to be totally off the wall to grab a knight from a4 and move it accross the board. It's not even an ambigious move situation.

I even checked my Mon Roi notation, because sometimes I've moved the wrong knight when it has been a matter of both knights being able to go to the same square.

In fact the very first tournament that I used the Mon Roi, I moved the wrong knight, and didn't notice it until 4 moves later when the position on the unit didn't match the board. I wasted 5 minutes trying to figure out where I went wrong!

Liquid Egg Product said...

Did you try using Chess Publisher v2? I'm not sure if it would make a difference--I really can't imagine they'd re-write the PGN parser for v2. It would be interesting to see, though.

Liquid Egg Product said...

Addendum: Chess Publisher v1 uses something called Chess Genie. I haven't bothered to ascertain the relationship between the two. Maybe v2 does use a different parser or something.

Blue Devil Knight said...

LOL--chesspublisher has invented a new variant of chess. Chesspublisher sure is quirky.

Polly said...

LEP: I haven't messed around with version 2. I don't like the way it scrolls, and I'm too HTML stupid to try you modifications to keep all the analysis nect to the main line. 1/2 the time when I insert a game, picture or diagram I get errors about incompleted tags. I don't even want to mess with trying to insert extra ones.

BDK: It's called Chess Publisher Random Chess. I wonder if we can market it. If Chess 360 can be castlinglong, I can specilaize in CP Random Chess on castlingqueenside. Though the way I've been playing maybe I'll call it "Polly Random Chess".

Glenn Wilson said...

The move is not ambiguous (except, perhaps to Chess Publisher, for reasons unknown). I just think making it explicit might help the software to work around its bug...

In fact the very first tournament that I used the Mon Roi, I moved the wrong knight, and didn't notice it until 4 moves later when the position on the unit didn't match the board. I wasted 5 minutes trying to figure out where I went wrong!

How does one correct that in MonRoi? Moving the pieces around until they match the position on the board in an edit mode? Backing up and re-entering the moves?

How can an opponent distinguish that from taking illegal advantage of trying out moves/positions on the MonRoi?

Liquid Egg Product said...

Can you post the pgn for the game? I'll give it a go.

Polly said...

LEP: Here is the PGN.

[Event "4 Rated Games!"]
[Site ""]
[Date "2007.11.29"]
[Round "2"]
[White "Polly Wright"]
[Black "Anatoly Tsinker"]
[Result "0-1"]
[WhiteElo "1700"]
[BlackElo "1972"]
[Source "MonRoi"]
[EventDate ""]
[TimeSpend "2894"]


1.c4 Nf6 2.Nc3 e6 3.g3 d5 4.cxd5 Nxd5 5.Bg2 c6 6.Nf3 Nd7 7.O-O Bd6 8.Ne4 Be7 9.d4 O-O 10.a3 Qc7 11.Qd3 e5 12.Neg5 N5f6 13.dxe5 Nxe5 14.Bf4 Nxd3 15.Bxc7 Nxb2 16.Rfb1 Na4 17.Be5 h6 18.Nh3 Ng4 19.Bf4 Bf6 20.Ra2 Nc3 21.Rbb2 Nxa2 22.Rxa2 Be6 23.Rd2 Rfd8 24.Ne1 g5 25.Rxd8+ Rxd8 26.f3 Bd4 27.e3 Bxe3+ 28.Bxe3 Nxe3 29.Nf2 f5 30.Ned3 Bc4 31.Ne5 Be2 32.h3 Kg7 33.f4 gxf4 34.gxf4 Kf6 35.Kh2 Nxg2 36.Kxg2 Rg8 0-1

Good luck!

Liquid Egg Product said...

It works fine in v2. It's not harder to use than v1, codewise, as long as you don't bother using my tinkering. :D

Anonymous said...

Yes, I experienced that before, also with a Chess Publisher choosing the wrong knight to jump to the wrong square.

See 13....Nh6 in this post.

Anonymous said...

The the Flash program on Chess Publisher is actually a Chessgenie. You'll find the (original) latest version on http://www.rootyhillchess.org or at its blogsite www.chessgenie.blogspot.com