Thursday, July 22, 2010

Shelby Lyman's Chess Column: Errors (item #1):

My fine local paper, the Trenton Times, has a daily chess column. The column doesn't say so, but Times employees assure me that this is Shelby Lyman's daily chess puzzle. Have you found errors in any of these daily chess puzzles? If so, please comment. I wish this entertaining puzzle column was more accurate.

I count Shelby Lyman as a friend, way back from 1959, and I enjoyed his many chess forays on TV. This column is usually fun, but it is also frustrating for the errors that appear in it. The quality of the column appears to change every few years, suggesting the possibility that various people have been hired to "ghost-write" the puzzles. (I have not contacted Lyman about this blog piece. I would be writing him instead of writing to all of you, if I knew how to contact his column.)

Part of my frustration comes from never seeing a correction. The column, years ago, rarely posted the worst possible error: giving the solution to some other problem than the one shown (and moving non-existent pieces). Perhaps six times a year there are mistakes: solutions that do not work, or alternate and better solutions. I wish the column published corrections when there were errors! Many of the errors could be avoided by testing the problems with computer programs specifically designed to test problems; I do not know why they seem not to be used.

Today, I made a determined deffort to find the puzzle in today's Lyman puzzle on the internet. Instead, I found other people who are frustrated that they cannot contact Lyman to discuss errors. This web page illustrates a catch by one 'Jedzz': I remember the diagram he shows, and -- gee -- the column must have more mistakes than I thought, because I did not catch this one.

In order to share Lyman's mistakes with you, I needed a program that would put chess diagrams in my blog. I am using Chess Diagrams, by DrAmbar (A Chatterjee). (Thanks!) I hope, I hope, my positions will be correct. I do not wish to insert any more errors. Today's error, the proverbial straw for me, is a minor one:


The caption implies that this position arose in the game Tregubov-Vorobiov (2001). The hint is: White to Play. Win a rook. And the hint is correct. I expected the solution to show several lines, but it only shows this (I added the question marks):
1. Qf7ch! Qd7? 2. Rc1ch?? Kd8 3. Qf8ch gets a rook.

Now if black had played 1 ... K moves, then 2. Qf8ch would indeed win a rook. But after 1... Qd7, white wins everything: 2. Rxb7ch Rxb7 3. Rxb7ch Kxb7 4. Qxd7ch K moves 5. Qg7 and black's remaining rook is trapped.

I will share one more Lyman puzzle error with you, and then: let's look forward, not back. Here's another position, in which "White Forces Mate" (Warning: Don't try to solve it!)


The given solution is: 1. Bf7ch Kh7 2. Bg6ch Kh6 3. Rh8ch Kg5 4. Rxh5 mate. And it would be mate, too, if black couldn't just play 4 ... Kg4. (It's possible that the position is incorrect; add a white pawn at h3 and it all works.)

Open Message to Shelby Lyman: Please check your problems by computer, and please print corrections! Thanks.

36 comments:

Kate Gasser said...

Shelby's column from today's Boston Globe (10/26/10) asks for White to win the queen. In fact, there is a mate in two.
white pieces: Q, e8; K, d2; R, e1; P, b4.
black pieces; Q, b6; K, d5; R, c7; P, f6.
Solution: 1. Qe4+, Kd6 (only); 2. Qe6 ++.

tobyr21@gmail.com said...

Kate, you're right. Thanks for saving me the trouble of posting that one. The final position is almost a traditional Epaulette mate.
- PB

Dave W. said...

The column from last Friday's SF Chronicle (10/22/10), which is the same as the one appearing here, has a hint of "Get a knight for a knight." The solution given is 1. ... Nxg3!; 2. Kxg3, Rg1ch!

But if instead 2. N(g5)-h3, then it seems to me that Black is left with a choice between 2 ... Rxh3; 3. Nxh3 (giving back the exchange for a pawn) or 2 ... Ra1; 3. Kxg3, Rxa2, giving up the knight for two pawns. Have I missed something here?

white pieces: Kg2; Nf4,g5; Ba3; Pa2,b3,c4,g3.
black pieces: Kc8; Rh1; Be5; Nf1; Pa7,b7,c6,f5.

tobyr21@gmail.com said...

Dave W,
The position is a bit complicated, but your move looks good to me.
Shelby should check his problems with a chess computer!
-PB

Anonymous said...

Dear bloggers,
I was trying to duplicate Shelby Lyman's list of moves for Volokitin and Maslak in the Columbus (Ohio) Dispatch. When I got to move 20, it could not be completed as it was listed. He must have copied the steps wrong from their game. Now on the internet I see others complaining that he makes numerous mistakes. I just crossed him off as a source for me to better my game.
Boots

tobyr21@gmail.com said...

If you get his daily column, do not give up hope. for the last few weeks, the daily problems have been more difficult and more accurate. It's possible that Lyman has someone helping him with the column, and he just got a better assistant.
- PB

timouthy said...

I found an alternate solution to the chess problem in yesterday's column (4/20/2011). The solution given by the paper works, but the one I came up with does also. I say move white queen to B5. This prevents black queen from checking white king and forces it to take white queen or scram. If black queen takes white queen, pawn takes black queen and threatens to become queen soon. If black queen scrams, you take black bishop. It is better if black trades queens because pawn at A4 will either become queen or black loses bishop to prevent it.

tobyr21@gmail.com said...

timouthy,
Plus ca change, plus c'est la meme chose.
-PB

Jim Carling said...

In the 5/22/11 St.Pete Times, the answer given is "mate in 2 - 1.Bd3ch! Nxd3. 2.Qf3 mate".
A simpler answer is Pf3 mate!

heykentswearingen-listen said...

Shelby Lyman's column is pathetically bad, even without the errors in the puzzles. He devotes an inordinate amount of space to hero worship of Bobby Fischer. When he does talk about current players, he never gives one of their actual games. Instead he gives a game score of some other players, with no commentary. By far the worst chess column I have ever read.

michael said...

Today's column in s f chronicle (6-9-11) has black's Q taking the h2 pawn for checkmate. But he gpawn is in the way now at h3! Does the Q jump over it?
Michael

tobyr21@gmail.com said...

Shelby should computer-check his problems! Actually, QxH3 (the legal move) does lead to mate, I think.
- PB

Anonymous said...

Lymans Main problem Ganguly v. Kunte, July 31, 2011 gives:

after Black plays 25. ... e5
Lyman remarks that White wins with
26. Nf5, stating "Black cannot stop 30. R(e)h3 (with the lethal threat of Rh8 mate)." But, 26. Nf5 is met with Bxf5. How does the R go to h3 with the Black B on the c8-h3 diagonal?

tobyr21@gmail.com said...

I saw that problem, and I had the same thought. I wonder if the position is incorrect, though I cannot see how to remedy it.
- PB

Anonymous said...

Chess by Lyman in today's SF chronicle is wrong, unless that white bishop at d1 is at f1. Otherwise Black would trade queens and pick up the black bishop.

tobyr21@gmail.com said...

I've been told that Lyman sends camera-ready content to the newspapers, so he is responsible for any errors. I wish he would acknowledge his occasional mistakes; it would make good copy. I also wish he would buy a chess program that specifically tests chess problems for errors!

I've known Shelby. He's a wonderful person, and I wish his column could leave a better reputation for him.
- PB

weaver said...

Today's column (1/12/2012) in the SF Chronicle has the solution for black to 'win the knight', which is great, except black could simply win the game instead.
The solution has for 3. Ke1 Bf2ch which would win the white knight, but as black, I'd prefer Bc3mate.

Jeffrey Pinyan said...

Today's column is about taking the rook... when Rxh7 is checkmate, no?

tobyr21@gmail.com said...

Jeffey Pinyan refers to the sunday feb 5, 2012 chess Column. Shelby Lyman's comment is: too easy for a Hint. (I must agree with that.) Lyman's solution is for white to move his knight, trapping black's rook. But J.P. is. Correct, Rxh7 is mate in one.

Let me repeat: I understand that Lyman ships camera-ready copy of his columns. It's a near-scandal that such a low-quality column, with no way to give feedback, is how chess is represented in many of our newspapers.
-PB

Jeffrey Pinyan said...

"It's a near-scandal that such a low-quality column, with no way to give feedback, is how chess is represented in many of our newspapers."

That's exactly what brought me, ultimately, to your blog today. I was looking for someplace to read / leave comments about the Chess column in my paper. All I had to go one was the author's name (and the text of the column).

Not making feedback possible is frustrating in this internet age!

Fred said...

I, too, am frustrated by the errors that frequently occur in the solutions provided to the "Chess by Lyman" columns that appear in the SF Chronicle. The one that appeared in the 2/5/12 paper is at the low end of the error scale. It says "1. Qxh5 does it. If 1. ... gxh5, 2. Bd1! (threatens the unstoppable. Bxh5 mate)." It's true, that would do it, but 2. Rg7 mate is quicker and easier. Can't the chess community do something thing about this? Can't someone persuade Lyman to ask 3 or 4 reasonably good chess players check his solutions before he has them published??

Fred said...
This comment has been removed by the author.
Fred said...
This comment has been removed by the author.
Fred said...
This comment has been removed by the author.
Fred said...

It looks like Lyman blew another one. The Chess by Lyman column in yesterday's (2-21-12) the San Francisco Chronicle had the board set up as shown at the following address:

http://www.flickr.com/x/t/0090009
/photos/75436879@N08/6775319210/

The Title is: BLACK HAS A CRUSHER with the HINT being Key is a knight fork.

The solution given is 1. ...Rf5ch! If 2. Qxf5, ... Nh4ch gets the queen.

But Rf5ch leaves an "out" for the king: 2. Ke2 which avoids the knight fork.

Better, IMHO, is 1. ...Re3ch, which leaves white one move: 2. Bxe3. Now Ne5ch defintely gets the queen.

tobyr21@gmail.com said...

Fred, there are two problems with your solution, I think. First:
1. ...Rf5ch
2. Ke2 Rxf2mate

Second:
1. ...Re3ch
2. Bxe3 Ne5ch
3. Ke4 Nxg4?
4. Rf8ch Kxf8
5. Bxg1 recovering the queen.

Still, I'm uncomfortable with Lyman's solution. I think that:
1. ...Qa1ch
2. Qg2 allows Nh4ch winning the queen without even giving up the rook. On any other move, 2.... Rxd2 probably leads to mate.
-PB

Jeffrey Pinyan said...

TPB, if white's "other move" after 1 ... Qa1ch is:

2. Kg3

And black decides to do

2. ... Rxd2

then there will indeed be a checkmate, but white will win:

3. Qc8ch Re8
4. Qxe8ch Nf8
5. Qxf8#

-----

Also, white can still threaten black's rook after 2. Qg2:

2. Qg2 Nh4ch
3. Kf4

-----

But it's all conjecture. ;)

Fred said...
This comment has been removed by the author.
Fred said...
This comment has been removed by the author.
Fred said...
This comment has been removed by the author.
Fred said...

TPB and Jeffrey Pinyan:

You are both right; my analysis did not go deep enough. I withdraw my critique of Lyman's 2/21/12 column; his solution is fine.

PS In your comments you suggest 1. .... Qa1ch, but you must mean 1. .... Qh1ch, right?

tobyr21@gmail.com said...

Fred, yes, I meant Qh1ch.

I'm kind of lazy, but I think I ought to give Shelby a call (I knew him in 1959) and ask him a few questions. Can anyone help me find his phone number? (Suggestions? email me: tobyr21@gmail.com)

tobyr21@gmail.com said...

I believe I have found shelby's phone number. More anon...
-PB

Anonymous said...

Shelby's "Beginner's Corner" for Saturday, February 4, 2012 printed a glaring mistake. He overlooked a simple mate-in-one solution (Rxh6++) instead indicating a "Nxd4!" to trade off a rook. Ironically for the "Hint & Explanation" he said "Too easy for a hint."

Here's the FEN position:
8/2R4p/4Nbp1/4Br1k/p4P2/P4K2/1P3P1P/8 w - - 0 1

Anonymous said...

Lyman's July 3 2012 chess puzzle has Qe8 for mate but black's queen is sitting on e2 for an uncontested capture of white's queen.

tobyr21@gmail.com said...

Yes. It's a pitiful error. I tried hard to solve that problem, and all the while I was thinking, is this another one of Lyman's typos? And it was.
- PB