Ask about Chess Problems

[Coyote]

Fine. In the spirit of friendly competition, I'm running this thread against i_h8's Anime thread to see which sinks to the bottom first. I consider this a perfectly valid entry in this forum, as I truly regard Chess Problems as an art form.

I may 'cheat' a little by posting a short discourse on some aspect of Problem Art as a means of bumping the thread...but then again I encourage i_h8 to do the same--just in the spirit of friendly competition of course!

[Gomez]

What's the easiest way to come up with good ones?

[Dr. Borodog]

This has nothing to do with chess "problems," per se, but I'm hoping you'll answer them anyway.

What are the top ten "book" openings? Why are they considered the best? What are the best defenses against them?



------------------
You will respect my philosophai.

[i_h8_evil_stuff]

[Coyote]

Gomez
What's the easiest way to come up with good ones?

Recall that I said I consider Chess Problems to be an Art Form. Your question is tantamount to asking what's the easiest way to paint a good picture. Nevertheless, I shall endeavor to give some semblance of an answer:

Study the works of previous composers--focus on the works that have been awarded prizes, but pay attention to the lesser known works as well. Understand just what ideas they were trying to put forth in these works, or all that follows will be for naught. Now go to the chessboard and try to work up some similar idea. Start simple--for example you might try to show three different mates given by pawns, or two mates allowed by unpinning a White piece. You'll fail miserably at first, but don't get discouraged! Go back to the works of those who've gone before and study how they avoided your failures. Now go back to the chessboard and gradually increase the complexity of the ideas you're working with until you have a fluent grasp of how all the pices interact with each other. Your final step--or perhaps this should be the first step--is to subscribe to several Chess Problem magazines to keep abreast of what's currently in vogue: you don't want to waste time working on an idea that went out of style 20 years ago!

Okay, so that sounds like a bit of work, but remember that you didn't ask me for an easy way, just for the one that's easiest.

Oh, and don't let the preceding discourage you--Chess Problem magazines are a bit demanding as to what they'll publish nowadays, but there's no reason why you can't compose just for fun--and an opportunity to post your puzzle in VSP.

Borodog
What are the top ten "book" openings? Why are they considered the best? What are the best defenses against them?

I'm sorry, but I'm really not qualified to answer this question, especially since the top ten tends to change from year to year, and it's been quite some time since I played competitively. You'd be better off asking some of the better players here such as Dragon Phoenix or borschevsky.

It's important to realize that while Chess Problems rely on the rules of chess as a framework, they are really a totally seperate entity; an art rather than a sport. As Pal Benko, a chess Grandmaster as well as Problem composer, once stated: "Chess Problems are like Tournament Chess in the same way as ballet is like wrestling."

[Agamemnon]

Not being a great/good/average/poor player like most on here, I have a chess problem question that will probably be laughed at but here goes....

In chess problems where problem states 'White checkmates in 3' or other examples, does One have to guess what moves black will make after your moves, or is it just a case of moving white 3 times to finish?

Go on, laugh now!

[mith]

Unless it's a helpmate, you have to account for all possible moves by black. Quite often, this is done by making a move that only gives black one or two options, but sometimes chess problems have several different lines of play.

In helpmates, you get to pick black's moves to help with the mating process.

[Chuck]

First, try sacrificing your most valuable piece. Chess puzzle composers like to require an apparently questionable move to throw solvers off.

[Vegetable]

Place 8 queens on a chessboard so that none of them attack eachother, then tell us whether or not it's possible.

[mathgrant]

Why are chess problems?

[MacadamiamaN]

http://www.milkandcookies.com/flash/num/8queens.swf (game that inspired veg's proposition)
NOTE: these people don't like you linking their stuff, so copy/paste that URL)
http://spaz.ca/aaron/SCS/queens/ (me cheating)
http://mn.glpics.com/8queens.jpg (me winning the game)



[This message has been edited by MacadamiamaN (edited 03-21-2004 06:50 PM).]

[extropalopakettle]

The "8 Queens" problem is a common exercise in computer programming courses. There are 92 solutions (counting rotations and reflections). There are 8^8 ways to place a queen in each column, but there are far more efficient ways than checking all 8^8 of those. Explanation and demo applet here: http://students.ceid.upatras.gr/~papagel/project/kef5_8.htm

[OcularGold]

so.. coyote. what are some chess probs you've written, and how'd you go about makin em?

[Coyote]

Aga
Why would I laugh? We all have our particular areas of ignorance. If I were to ask you some painfully naive question about opera, would you laugh at me or try to instruct me? (Ok, I'll admit it's possible you might do both.)

mith answered the question pretty well (thanks, mith!), so I'll just add this little bit: there really are chess problems where White simply keeps moving until he checkmates Black--in the required number of moves, of course! Such problems are known as Series-movers, or in this specific case, Series-mate. (Other types in this category include Series-helpmate, Series-selfmate, Series-reflex-mate, etc.)
Even granting this huge advantage to White doesn't always make the goal easily achieved:
G. Sphicas, 1987
Chess Board

---

Player 1 Player 2

---

Series-mate in 41
(White plays 41 consecutive moves, checkmating black on the final move. One restriction is that White may not place himself or Black in check during the course of the solution)

Chuck
I know your comment wasn't actually a question, but I'd like to address it anyway, as it points out a common misconception about chess problems. It's considered poor form to have the opening move (or Key-move) be an overly strong or aggressive one, which is why you rarely see problems that commence with a check or capture. A corollary to this principle is that especially weak-looking keys are considered more 'artistic', and composers will try to include such a move if it can be done without compromising the actual idea behind the problem. However, 'fooling' the solver is not usually the main purpose behind such moves, and in fact sometimes an especially strong move has to be resorted to if the composer is trying for an especially difficult idea. I've actually seen problems that start with 1.PxQ(Q)!

Of course, this doesn't mean that sacrificing a piece can't be made into the central idea of a problem:

M. Marble, 1909
Chess Board

---

Player 1 Player 2

---

Mate in Two

Vegetable
Others here have answered your request. Actually, this and similar puzzles such as knight-tours are more properly considered mathematical recreations than actual chess problems. However, there is a sub-genre known as Construction Tasks in which composers will try to set or break certain records such as greatest number of possible checkmates, longest series of consecutive checks, &etc.

mathgrant
I think a better question would be 'Why are mathgrant?

Ocular Gold
Sorry, I'll have to address your question after work, as I've nearly run out of time at the moment.

Solutions to problems above, in inv:

G. Sphicas:
To save space, only captures and the final move of the WK are given.
2.Kxd6 8.Kxc1 15.Kxh3 17.Kxf1 26.Kxd1 28.Kxf3 30.Kxg1 36.Kxa7 40.Kf4 41.Qxg8#

M. Marble:
1.Be4
1...Pxe4 2.Qd8#
1...gNxe4 2.Ne6#
1...fNxe4 2.hRd3#
1...fRxe4 2.Nf5#
1...eRxe4 2.Rc1#
1...Qxe4 2.Rc2#
1...Kxe4 2.Rc4#

[OcularGold]

oh, and a couple more questions:

what is the first documented chess problems? when did helpmates and series mates get "invented"?

are there different styles of chess problems as there are different styles of paintings/musical pieces?

thanks for your time, coyote

[Agamemnon]

In said chess problems, are there several solutions or just the one?

[Coyote]

Returning to OcularGold's question:

I've had about a hundred chess problems published, of which most of my favorites have either already been posted in VSP or dealt with odd chess variants I didn't consider suitable for that forum. Many of the former ones have unfortunately been pruned, but you can still see some here and here. Here's one that hasn't appeared yet. This is the first problem I ever got published in a periodical devoted strictly to chess problems. It's also possibly one of my best works, which means it's all been downhill from that point.

Chess Board

---

Player 1 Player 2

---

Helpmate in Four
(a) Diagram
(b) With c4N on d4 instead
(c) With c4N on d6 instead

(Helpmate: Black and White cooperate to reach a position where Black is checkmated. Unlike regular chess problems, in helpmates Black moves first. I have no idea why this came to be the standard convention.)

It's difficult for me to verbalize the process I go through when I compose, but here's an attempt, taking the first problem in the Leftovers thread as an example.

I first began with an Idea to illustrate: I thought it would be amusing to do a two-mover where Black could give check either by castling or by simply moving the rook to d8. Obviously, the WK would have to be on the d-file, and the checks would have to be answered by having a piece move to block the check, and at the same time discover checkmate. At this point I had something like this to work with:
Chess Board

---

Player 1 Player 2

---

This is a very rough attempt to illustrate the nebulous image in my mind at that point. In reality I just imagined the king and rooks as 'somewhere on their respective files' and the bishops as 'somewhere along their diagonals'. The pawns represent 'some random piece that discovers mate'. Nothing was actually on any specific square though.

I quickly dismissed the bishops as being too awkward to work with, which left the two rooks. Obviously I'd need a knight or bishop to discover mate, and by placing them on c5 and e5 they could also take part in guarding some of the squares around the BK. The big breakthrough came when I realized that by using knights I could move the WK all the way up to d6, thus guarding the squares on the 7th rank. Things were starting to shape up pretty good now:
Chess Board

---

Player 1 Player 2

---

Time to start doing some 'detail work'. On the left side of the board I needed to deal with the b7 square (which would be unguarded after 1...0-0-0+ 2.cNd7), and that unwanted check 1...Ra6+. My first thought was WPa6, BPa7; but at this point it began to dawn on me that I had a little retro-analytic problem to deal with as well--strictly speaking, 1...0-0-0 wasn't even available as a move, since either the BK or BR must have made the previous move!

Okay, so put Black pawns on a6 and b7. It's usually considered better form to add Black units to deal with constructional matters anyway. Now for the right-hand side of the board, where I had to guard f7 and deal with 1...Kf8. A WB on a2 would not only deal with the first issue, it would also allow 2.eNd7 to answer the second issue, if I added a BPg7.

Now I needed a threat a of some sort, to motivate the two checking defenses, and to answer such random moves as 1...Pa5. Since both d8 and f8 were unguarded, it would clearly have to be something along the lines of 2.Rh8#--but I already had both rooks busy with other matters! Worse yet, 2.Rh8 would also answer 1...Kf8. Alright, so get rid of the WBa2, move BPg7 to f7, use WQ somewhere along the g-file to threaten 2.Qg8#
Chess Board

---

Player 1 Player 2

---

Okay, that works, but it's not too satisfying from an artistic standpoint. It's really bad form to use the queen in such a minor role, and as basically a 'third rook'. She should take a more active role, and employ her diagonal powers as well. Time to polish the position a little.

That's when another flash of inspiration hit me. If I replaced the c1R with the queen, I could place her on c4 and give her diagonal acess to g8. Just replace the BPf7 with WPg6 and...wait a minute! I didn't even need a pawn over there any more, as 1...Kf8 could be answered with 2.Qf7! Sometimes good fortune plays an important role in composing problems.

Nothing left now but to find a Key-move for White. Simply starting the queen on some other square for 1.Qc4 would work, but you should always look for something better. I noticed that by starting the WK on e6 I could not only uncover the queens attack on f8, but also have the Key-move allow the Black checks--a very good, thematic Key. And now I'm done.

Well, not quite. This problem was also selected to show the problem-composers greatest nemesis--that of someone else having discovered your idea earlier (this is known as an anticipation). I composed this problem twenty years ago, but it wasn't until a few months ago that I came across the following:

A. C. White, 1918
Chess Board

---

Player 1 Player 2

---

Mate in Two

So my problem had already been thought up by someone else--nearly 70 years earlier! I had a couple of small consolations here: Alain White was a composer of considerable skill in his day, so the fact my efforts had duplicated his showed that my constructional technique had been good. And he had neglected the retroanalysis bit, so in his problem 0-0 wasn't strictly possible. Still, this has to be considered a complete anticipation, alas.

And that's all I have time for now, as Ocular's next question deserves a lengthy reply. I'm sorry I'm being a bit slow on answering, but after today I have two days off work, and can hopefully catch up a bit.

[Antrax]

This was a fascinating read.
Antrax

------------------
"Look, that's why there's rules, understand? So that you think before you break 'em" - Lu-Tze, Thief of Time

[This message has been edited by Antrax (edited 03-23-2004 07:36 AM).]

[Kd]

*raises hand* Uh, yeah. "En passant" - that's a proper, non-cheating move, right? (The thing with the pawn, etc.) So... why won't computer chess programs let me do it?

[extropalopakettle]

Are you sure you know the rules about when you can do it? It doesn't come up very often. Some programs allow it.

http://www.conservativebookstore.com/chess/enpass.htm

[This message has been edited by extropalopakettle (edited 03-23-2004 01:37 PM).]

[Kd]

Oh, ok... *curses whoever taught me chess* Thanks for that, extro.

[Chuck]

My favorite move is forward castling. My king has never moved and I promote a pawn to a rook on the king's eighth rank. Since this rook has never moved I can use it for castling, moving my king forward two squares and putting the rook immediately behind it.

I thought about making a chess problem that requires this move in its solution but people would probably just yell at me.

[Coyote]

Back to Ocular Gold's questions:

quote:

---

what is the first documented chess problems? when did helpmates and series mates get "invented"?

are there different styles of chess problems as there are different styles of paintings/musical pieces?

---

The first known chess problems appear in Arab manuscripts dating back to the 9th Century AD. Typically, they had White, in an apparently hopeless position and menaced with immediate mate next move, saving himself thanks to a series of piece-sacrifices. Chuck would have loved them. They were often accompanied by little stories explaining how the situation came about. One of the better known ones is the story of Dilaram, which is also interesting as possibly the first recorded case of a kibbitzer butting in on a game in progress:

quote:

---

Ancient Arabic written over a thousand years ago reveal the story of the nobleman Murwadi, who had a wife (one of many!) named Dilaram (literally translated as the heart’s ease). One day Murwadi was playing an ancient form of chess known as ‘shatranj’ and he played for some stakes. While playing, the stakes grew higher and higher until Murwadi ultimately staked his wife Dilaram on the game. Unfortunately, the game didn’t go too well for him and defeat seemed inevitable as the game came to the position shown below. Now, in the present position, Murwadi, who played the white pieces, was to move, expecting to be mated on his opponents return move.

Dilaram however saw how the happiness of her life could be saved, and shouted to her husband "Sacrifice your rooks, but not me!"

---

Chess Board

---

Player 1 Player 2

---

The 'bishop' in the diagram is actually the ancient shatranj piece called the Alfil, which leaps, knightlike, to squares two diagonal moves away (so in the diagram, the Alfil can move to either f1 or f5. The solution given was 1.Rh8+ KxR 2.Af5+ Kg8 3.Rh8+ KxR 4.Pg7+ Kg8 5.Nh6#, which isn't strictly accurate, since Black can delay checkmate by interposing his b2R, but you get the general idea.

Chess problems in their more modern form date back to the mid-1800s, when most of the strategic ideas known today first began to be discovered and given serious treatment, starting with Loveday's famous 'Indian' problem in 1845. Not that there hadn't been plenty of problems published before that date, but they generally took the old form of 'White makes a string of sacrifices to win'.

Sam Loyd is widely believed to have invented the Helpmate, in this problem published in 1860:
Chess Board

---

Player 1 Player 2

---

Helpmate in Three
Solution: 1.Kf6 Ra8 2.Kg7 Bb8 3.Kh8 Be5#

The idea wasn't exactly an overnight success--over the next fifty years only a scant handful of helpmates were published, until the use of such things as twinning and multiple solutions allowed composers to add a little more thematic content to such problems.

I'm not really sure when the first series-mover appeared, but I suspect they were introduced by T. R. Dawson sometime in the 1940's. Dawson, who is often referred to as 'The father of fairy-chess', introduced a great number of variant-chess pieces and conditions during that period. At any rate, the earliest series-mover I have knowledge of is one that he published in 1947.

--------------------------

There are indeed many different styles of chess problems--more, in fact, than I can really fully describe here. But here's a few of the major divisions:

In some problems, the focus is on producing interesting mating positions, most often model mates. For a mate to be considered a model mate, all White pieces (except king and pawns) must take part, and no square around the Black king may be double-guarded, or both guarded and blocked by a Black piece. Such problems often feature echoed mates, which means that a particular mating position is duplicated in a different location. Since my helpmate in four posted above illustrates this idea, and I forgot to include the solution, I'll list it here as an illustrative example:

(a) 1.Kd5 Kc2 2.Kd4 Pd3 3.Ne5+ Kd2 4.Qd5 Pe3#
Take note of the final position, especially the positions of the different pieces relative to each other.

(b) 1.Qd6 Ke3 2.Ne6 Pd4 3.Kd5 Kd3 4.Rc6 Pe4#
The mating position has been exactly duplicated, with everything moved up one square.

(c) 1.Rf5 Pe3 2.Kd5 Pe4+ 3.Ke5 Ke3 4.Rf6 Pd3#
The position has not only been moved up, but also reflected. Echoed mates can also be rotated, though I don't think I could have managed it in this case.

Other problems focus on strategic play, such as interferences, clearances, ambushes, &etc. This style of composing covers such a vast range of different types of problems it defies efforts to find a single illustrative example, so here's one chosen at random. Try to keep track of the complex interplay of cross-checks, pinning, and unpinning here:

G. Bouma, 1965
Chess Board

---

Player 1 Player 2

---

Mate in Three
Solution: 1.Kb8 (threatening 2.Qe5+ and 2.Nxe7+)
1...Bg3+ 2.Ne5+ Rd6 3.Nxg4#
1...Rb1+ 2.Nb4+ Bb6 3.Nd5#

(I've been going back and forth on the issue of whether to put solutions in {invisible} or not. On the one hand, the purpose of this thread is to instruct rather than post puzzles for solving, so I'd prefer to just keep the answers visible, since playing through the solution is the whole point of putting up illustrative problems. On the other hand there's always a chance some people would prefer to solve them on their own. If anyone wants me to start using {inv} again, let me know)

Yet another type of problem centers on changed play, that is, Whites response to a particular Black move is different depending on whether it occurs in set play (play that would happen if we assumed Black were to move first), virtual play (play following a 'try', which is a move that almost solves the problem, but is defeated by as single Black move), and actual play (the play that follows the actual key move). This was once a strictly two-move idea, but recently composers have been applying it to longer problems as well. Often the different phases are tied together by a common idea, as in the following, where all the replies center around Black interfering with the influence of his own rooks:

O. Stocchi, 1949
Chess Board

---

Player 1 Player 2

---

Mate in Two

code:

---

Set Play: 1.... Pg6 2.dNe6# 1...Bb2 2.Nb3# 


Try Play: 1.Nf5 Pg6 2.Bd6#  1...Bb2 2.Bb6#  but 1...Rb4! defeats


Key Play: 1.Nc2 Pg6 2.Qxc6# 1...Bb2 2.Qb4#


  (Threat in both Try and Key is 2.Qd4#)

---

And I have to mention one of my favorite types of problem, the Logical Problem. In a logical problem, white has a 'mainplan' that threatens mate, but Black has an adequate defense to this plan. So White first plays a 'foreplan', which threatens a different mate. Black also has an adequate defense to the foreplan, but employing it screws up the defense to the original mainplan, allowing White to return to it.

The interesting bit is that often the foreplan will itself need a foreplan before it can be used successfully--in fact sometimes a whole string of foreplans are needed, producing an effect similar to knocking over a chain of dominoes:

H. Lepuschutz, 1936
Chess Board

---

Player 1 Player 2

---

Mate in Six

White's mainplan is to play 1.Rb6 followed by 2.Rxb5#, but at the moment this can be defeated by 1...Pb1Q. So he tries the foreplan 1.Rd6 (threat 2.Rd5#), and if Black defends with 1...Bb3, he can return to the mainplan with 2.Rb6. But if Black defends with 1...Bf3 instead, the foreplan won't work, so we need a foreplan for the foreplan. (Got that?)

So let's try 1.Re6 (threat 2.Re5#). Now after 1...Nf3 2.Rd6, 2...Bf3 Is no longer possible and Black is forced to play 2.Bb3. But this still doesn't work, because Black can play 1.Pf3 instead! Time for yet another foreplan. After 1.Rf6 (threat 2.Rf5#), 1...Ng3 will eliminate 2...Pf3 as a defense after 2.Re6. But...you see where this is going, don't you? That's right, 1.Rf6 also fails after Black chooses 1...Bg4 as his defense. So one more foreplan, and this time the final one.

1.Rg6 (threat 2.Rxg5#). Now watch the dominoes fall!
1...Rg4
2.Rf6 Ng3
3.Re6 Nf3
4.Rd6 Bb3
5.Rb6
And White mates next move.

Whew!

Okay, Aga. I'll get to your question next, as soon as I work the kinks out of my fingers.

[OcularGold]

thanks coyote! that was a really good read.

oh and personally, having the answers not invised is good for me, but either way is fine. reading through the explanation for that last one was cool.

[mathgrant]

I have a chess problem. What's the best way to make sure it's half-decent and not 100% bad and stuff?

I confess, I posted mostly because I wanted chess to beat anime.

[This message has been edited by mathgrant (edited 04-01-2004 01:14 PM).]

[Sniklac16]

The only ches problem I have right now is trying to find someone who wants to play me.

------------------
"Stories just show the writers imagination but poetry is writing from the heart"

[i_h8_evil_stuff]

I've just got one quick Chess problem question:
You thought you'd win the race, didn't you?
I can't seem to figure this one out.
Only three from the bottom before I bumped this.
Black to play and mate in 2.
Chess Board

---

Player 1 Player 2

---

[Dread Pirate Westley]

I think you need to move everything one square left.
It was a good bump, though.
Otherwise, White can do b1, then escape to a2...Then get mated when he's forced to a3.

[i_h8_evil_stuff]

Hm. Well, I'll need to tell my friend that he was mistaken in borad position, then.

Chess Board

---

Player 1 Player 2

---

Just making sure. From here, it's 1. ... Re1 2. Ka1 Rxc1#, right?

[borschevsky]

Yes, that's right.

Incidentally, this is the reason that you defend a B vs R endgame by putting your king in the corner that's the opposite colour as your bishop.
Chess Board

---

Player 1 Player 2

---

Rb1 here is stalemate, and if black ever tries to mate along the h-file, white just puts the bishop on h2 instead of g1.