Hi guys,
Do any of you guys know actual chess games or problems where queening
a pawn to a queen instead to just a rook or a bishop gives results
worse than queening to rook or bishop?

The reason I'm asking for these is I was having a discussion about
"optimizing" a computer chess engine with an expert in the field
(Guess who that is.). I first suggested that the chess engine could
just skip over the analysis of lines where a pawn is queened to a
bishop or a rook since queening to a queen should give the same, if
not better results. Lines where the pawn is queening to a knight
should not be skipped over, of course, because its movement is not
encapsulated in that of a queen. But a few hours after I got a
response from the expert, an abstract exception popped up in my head.
That is, queening to a queen could actually be worse than queening to
just bishop or rook.

Now I need actually example game setups that can confirm my fears.

Kal

"Seabook" wrote in message
om...
Hi guys,
Do any of you guys know actual chess games or problems where queening
a pawn to a queen instead to just a rook or a bishop gives results
worse than queening to rook or bishop?

The reason I'm asking for these is I was having a discussion about
"optimizing" a computer chess engine with an expert in the field
(Guess who that is.). I first suggested that the chess engine could
just skip over the analysis of lines where a pawn is queened to a
bishop or a rook since queening to a queen should give the same, if
not better results. Lines where the pawn is queening to a knight
should not be skipped over, of course, because its movement is not
encapsulated in that of a queen. But a few hours after I got a
response from the expert, an abstract exception popped up in my head.
That is, queening to a queen could actually be worse than queening to
just bishop or rook.

Now I need actually example game setups that can confirm my fears.

Kal
Many,many examples. If you really want to blow your mind, have a look at the
"Babson Task" articles on Tim Krabbes site,
http://www.xs4all.nl/~timkr/chess2/diary.htm where there are compositions
with different promotions from the same position..

"Seabook" wrote in message
om...
Hi guys,
Do any of you guys know actual chess games or problems where queening
a pawn to a queen instead to just a rook or a bishop gives results
worse than queening to rook or bishop?

The reason I'm asking for these is I was having a discussion about
"optimizing" a computer chess engine with an expert in the field
(Guess who that is.). I first suggested that the chess engine could
just skip over the analysis of lines where a pawn is queened to a
bishop or a rook since queening to a queen should give the same, if
not better results. Lines where the pawn is queening to a knight
should not be skipped over, of course, because its movement is not
encapsulated in that of a queen. But a few hours after I got a
response from the expert, an abstract exception popped up in my head.
That is, queening to a queen could actually be worse than queening to
just bishop or rook.

Now I need actually example game setups that can confirm my fears.

You can search a large PGN files for "=R" or "8R" (or "=N", or "8N", etc.)
patterns. For example:

[Event "?"]
[Site "Vienna gt"]
[Date "1904.??.??"]
[Round "?"]
[White "Schlechter C"]
[Black "Wolf H"]
[Result "1-0"]
[ECO "C30/03"]

1.e4 e5 2.f4 Bc5 3.Qf3 d6 4.fxe5 dxe5
5.Qg3 Qf6 6.Nf3 Nd7 7.Bc4 Qg6 8.Qxg6 hxg6
9.Nc3 c6 10.Ng5 Nh6 11.d3 Nb6 12.Bb3 Ke7
13.Nf3 Nd7 14.Nd1 f6 15.Be3 Bxe3 16.Nxe3 Nc5
17.Nh4 Nxb3 18.Nxg6+ Kd8 19.axb3 Re8 20.O-O Be6
21.h3 Bf7 22.Nh4 Ke7 23.Nhf5+ Nxf5 24.Nxf5+ Kf8
25.h4 Be6 26.g4 Bxf5 27.Rxf5 Ke7 28.Kg2 Rh8
29.Kg3 a5 30.g5 b6 31.Kg4 Ke6 32.Raf1 Rag8
33.h5 Ke7 34.R1f2 Kd7 35.R5f3 Ke7 36.Rf1 Kd7
37.Kh4 Ke6 38.gxf6 g6 39.f7+ Rxh5+ 40.Kg4 Rf8
41.Rf6+ Kd7 42.R1f2 Rh1 43.Kg5 Rfh8 44.f8N+ 1-0

first, it's called ' pawn promotion', not 'queening'. it's queening iff you
make a queen, so speak english.
Second,crack open capablanca's book for chess beginners. It's called "chess
fundamentals', there you will find examples of pawn promotion to queen
leading to stalemate when a promotion to a rook wins.
Third, learn to play chess before you advise experts on how to play.
"Seabook" wrote in message
om...
Hi guys,
Do any of you guys know actual chess games or problems where queening
a pawn to a queen instead to just a rook or a bishop gives results
worse than queening to rook or bishop?

The reason I'm asking for these is I was having a discussion about
"optimizing" a computer chess engine with an expert in the field
(Guess who that is.). I first suggested that the chess engine could
just skip over the analysis of lines where a pawn is queened to a
bishop or a rook since queening to a queen should give the same, if
not better results. Lines where the pawn is queening to a knight
should not be skipped over, of course, because its movement is not
encapsulated in that of a queen. But a few hours after I got a
response from the expert, an abstract exception popped up in my head.
That is, queening to a queen could actually be worse than queening to
just bishop or rook.

Now I need actually example game setups that can confirm my fears.

Kal

If you look this subject up in a book, refer to "under promotion."

Any smothers mate where the knight promotion works and the queen doesn't would
be an example of what you're talking about. It would be a little more difficult
to find a situation where a promotion to a bishop or a rook would be better was
the queen combines the moves of those two pieces. If the mate wasn't on the
move, then it is easier to do.

On 2 Oct 2003 17:59:08 -0700, (Seabook) wrote:

Hi guys,
Do any of you guys know actual chess games or problems where queening
a pawn to a queen instead to just a rook or a bishop gives results
worse than queening to rook or bishop?

The reason I'm asking for these is I was having a discussion about
"optimizing" a computer chess engine with an expert in the field
(Guess who that is.). I first suggested that the chess engine could
just skip over the analysis of lines where a pawn is queened to a
bishop or a rook since queening to a queen should give the same, if
not better results. Lines where the pawn is queening to a knight
should not be skipped over, of course, because its movement is not
encapsulated in that of a queen. But a few hours after I got a
response from the expert, an abstract exception popped up in my head.
That is, queening to a queen could actually be worse than queening to
just bishop or rook.

Now I need actually example game setups that can confirm my fears.

Kal

I don't know any positions from games, but i have seen a few
compositions - the only one i can remember off the top of my head is a
classic:

8/8/1kp5/3R4/8/8/8/K7 b

black to play and win:

.... c7
Rd6+ Kb5
Rd5+ Kb4
Rd4+ Kb3
Rd3+ Kc2
Rd4! c8=R!! {c8=Q draws after Rc4+! Qxc4 stalemate}
Ra4 Kb3
0-1

The Polgar 5334 tactics book has a number of mate in 2 or 3 problems
where underpromotion to R or B are necessary to avoid stalemate, but i
don't much feel like searching them down.

"marc margolies" wrote in message ...
first, it's called ' pawn promotion', not 'queening'.

First, it's called: 'pawn promotion', not: 'queening'.

it's queening iff you
make a queen, so speak english.

It's queening only if you promote to a queen, so speak English.

Second,crack open capablanca's book for chess beginners.
It's called "chess
fundamentals',

Second, crack open Capablanca's book for beginners, it's called:
"Chess Fundamentals".

there you will find examples of pawn promotion to queen
leading to stalemate when a promotion to a rook wins.

Therein, you will find examples of pawns promoting to a queen and
creating stalemate positions, whereas underpromotion to a rook from
the same position, wins.

Third, learn to play chess before you advise experts on how to play.

Third, learn to speak and write clearly in English before you advise
others: "...so speak English.".

Sorry, Marc, I couldn't resist....

"Seabook" wrote in message
om...
Hi guys,
Do any of you guys know actual chess games or problems where queening
a pawn to a queen instead to just a rook or a bishop gives results
worse than queening to rook or bishop?

The reason I'm asking for these is I was having a discussion about
"optimizing" a computer chess engine with an expert in the field
(Guess who that is.). I first suggested that the chess engine could
just skip over the analysis of lines where a pawn is queened to a
bishop or a rook since queening to a queen should give the same, if
not better results. Lines where the pawn is queening to a knight
should not be skipped over, of course, because its movement is not
encapsulated in that of a queen. But a few hours after I got a
response from the expert, an abstract exception popped up in my head.
That is, queening to a queen could actually be worse than queening to
just bishop or rook.

Now I need actually example game setups that can confirm my fears.

Kal

Apart from AVOIDING stalemate, there is a second possible reason for
underpromotion to rook or bishop, as I learned from a recent article by Jan
Timman in New in Chess Magazine. He gives several studies in which
underpromotion on the contrary serves to BRING ABOUT a stalemate.

Here's a simple example, a study by Archakov and Zinar from 1985:
8/1p4P1/1P6/KP2p3/1Pkppp2/5ppp/1P6/8 w - - 0 0
White to play and draw.

1.g8B+!
After 1.g8Q+? Black's pawns (he still has all 8 of them!) would prevail.
1...Kd3 2.Bb3
and Black cannot prevent White from stalemating himself with 3.Ba4 and 4.b3.

For more examples see Timman's article in New in Chess 2003/5, page 93.

Tobi

Joe Schoeman wrote:

Many,many examples. If you really want to blow your mind, have a look at the
"Babson Task" articles on Tim Krabbes site,

Those are not quite relevant -- the Babson task is to show all possible
promotions on one side, answered by the same promotion on the other, not that
minor promotions are better than 'major' ones.

Here is one example, fairly well known to beginners, that the obvious
promotion is the wrong one:


E. B. Cook: 8/6P1/8/8/8/8/2K5/k7 (Mate in two).


And here's one where neither Queen not Rook will do:

Sam Loyd: 8/PP3k2/5P2/5K2/8/8/8/8 (Mate in three)


You wouldn't want a chess program to go wrong in this
simple situation, no matter how contrived it is.

--
Anders Thulin http://www.algonet.se/~ath

your stylistic advisement is noted. yet, did you any content to my message?
also, 'pawn promotion' is only refered to as 'underpromotion'
problematically as a theme. by changing my term you are begging the
question, Mark. so clearly one of your 'improvements' does not appear to
improve the clarity of my phrasing. 'iff ' is a standard abbreviation for
'if and only if.' since those relative clauses which you added to my
sentences did not make them more understood, they were unnecessary. were i
your teacher, i'd remove them.
but i am glad to be read by a thinking person even if he doe not like my
style or attitude. thanks.

"Mark Houlsby" wrote in message
m...
"marc margolies" wrote in message
...
first, it's called ' pawn promotion', not 'queening'.

First, it's called: 'pawn promotion', not: 'queening'.

it's queening iff you
make a queen, so speak english.

It's queening only if you promote to a queen, so speak English.

Second,crack open capablanca's book for chess beginners.
It's called "chess
fundamentals',

Second, crack open Capablanca's book for beginners, it's called:
"Chess Fundamentals".

there you will find examples of pawn promotion to queen
leading to stalemate when a promotion to a rook wins.

Therein, you will find examples of pawns promoting to a queen and
creating stalemate positions, whereas underpromotion to a rook from
the same position, wins.

Third, learn to play chess before you advise experts on how to play.

Third, learn to speak and write clearly in English before you advise
others: "...so speak English.".

Sorry, Marc, I couldn't resist....

"Seabook" wrote in message
om...
Hi guys,
Do any of you guys know actual chess games or problems where queening
a pawn to a queen instead to just a rook or a bishop gives results
worse than queening to rook or bishop?

The reason I'm asking for these is I was having a discussion about
"optimizing" a computer chess engine with an expert in the field
(Guess who that is.). I first suggested that the chess engine could
just skip over the analysis of lines where a pawn is queened to a
bishop or a rook since queening to a queen should give the same, if
not better results. Lines where the pawn is queening to a knight
should not be skipped over, of course, because its movement is not
encapsulated in that of a queen. But a few hours after I got a
response from the expert, an abstract exception popped up in my head.
That is, queening to a queen could actually be worse than queening to
just bishop or rook.

Now I need actually example game setups that can confirm my fears.

Kal