Until recently, my aim in breakthrough was not to lose any games anymore. It's difficult since there is no draw. So, I'd have to win all my games in order no to lose.
I don't particularly like to win.
But losing is harsh and humiliating. If you lose with both colors, you know you were wrong somewhere.

In that sense, Chess is less tough.
Mistakes are irrelevant: it is always equal.
Now, if you insist in playing badly, you may eventually lose.

Anyway, since some players (wanderer_c, Stop Sign, gardiyan, michelwav, luffy_bot…) seem to be against the idea of me no more losing games, I concede. You win!

No, my main goal is to share with you the beauty of breakthrough.
(Actually, it sounds like a nobler purpose than the latter.)

This thread will be devoted to breakthrough puzzles (hard, subtle and/or beautiful problems).

Let this be the first one.
What happens if White plays 47.g3-h4?
It is a *** (3 stars) problem.
(\* being easy and **** being very complex)

http://www.littlegolem.net/jsp/game/game.jsp?gid=1606181

Some indication: (**slight spoiler**)
For problemists, beauty can come from an exercise in style on a theme. Here the theme is:
If White plays A, Black wins with X (and lose with Y)
If White plays B, Black wins with Y (and lose with X)
This theme is one of the simplest and could be implemented with trivial examples.
Here, the variations are interesting and the apparition of the theme is subtle.
Have fun!

In each game lost. such a case is very rare to lose. I think the important thing is to be the best and win against the best.

I love this game very much, I think the breakthrough is an art. This game is very fun competing against a powerful, much more enjoyable if you win.

I imagine. plays a white wins …. 47.g3-h4 or g3-f4

good moves (edbonnet & wanderer_c)

[game;id:1606181;title:What happens if White plays 47.g3-h4?]

@Carroll: That helps. Thanks!

@basat: After a very quick look, Wanderer does not seem to like your move(s) but prefers c5-d4 – but I don't have time right now to try to understand why. I'll try to look at it more tonight and see if I understand why and/or understand edbonnet's hints and comments.

On 47.g3-h4, Black wins with 48.c5-d4.

If 49. exd4 50.e5-f4! -+

Now, there is two ways to defend e3: d1-d2 or c1-d2. Both create a crucial weakness, respectively e2 and e3.

If 49.d1-d2 50.e6xf5 51.e4xf5 52.e5-e4! 53.d3xe4 54.d4-d3 55.c1-c2 56. d3-e2! -+
If 49.c1-d2 50.d4xe3 51.d2xe3 52.e5-f4!! 53.e3xf4 54. e6-e5 55.f4xe5 56.f6xe5 57.h5-g6 58.e5-f4 (threatening e3 and winning a decisive tempo) 59.d1-d2 60.f4-f3 -+

Note that Black has to be particularly accurate.
For instance, taking on f5 first, loses: 48.e6xf5 49.e4xf5 50.c5-d4 51.c1-d2! +-
Moving the g4 pawn first, also loses: 48.g4-h3 49.h5-g6 50.c5-d4 51.d1-d2! +-
with a nice resurgence of the theme.

@Carroll: Thanks! How do you include diagrams? The diagram “edbonnet-wanderer_c” you've included seems to be dynamic (the position is updated while the game goes on) instead of static.

In this position, from the present championship 1,

http://www.littlegolem.net/jsp/game/game.jsp?gid=1568417&nmove=56

Kyle decided to take on d5.
And Halladba resigns a few moves later in a winning position.
Can you spot a win for White at move 57, instead of 57.c4xd5? ?
This is a ** problem.

[game;id:1606181;move:46;title: *** What happens if White plays 47.g3-h4?]

[game;id:1568417;move:56; title: ** White to play and win.]

You include positions like this:
[ game;id:123456] or [ game;id:123456;move:20] or [ game;id:123456;move:20;title:some text]. For tournament table you can write [ tournament;id:chess.ch.10.2.1].

As I did not specify move number, this points to last move.
You can click on picture to get into the game.

Haha you beat me into it!

In this position, Ray Garrisson blundered with 57.e3-e4?? and eventually lost the game.
What would you play instead?

[game;id:1318373;move:56;title:* White to play and win.]

57.f3-e4 58… Ray Garrisson win!

You are way too strong for * (one star) problem ;)
What about the other two?

*57.f3-e4* 58.d5xe4 (white outdoes all other moves. it is futile to even move his black …) 59.d3xe4 and white somehow find a way out of the square f5. (1-0)

My original intention was to play 57.f3-e4, but as I wrote to kyle during the game, I returned from vacation and mistakenly played the fingerfehler which I immediately regretted….Kyle showed no mercy!

It seems that nobody wants to suggest an answer in the kyle douglas-halladba game.

The simplest way to victory in the position of the diagram is the 'brute-force':

57.c5xd6 58.c7xd6 59.d3-d4 (threatening c5) 60.a7-b6 61.d4xe5 62.d6xe5 63.e3-d4!
64.e5xd4 65.e4-e5+-. Black “swiss cheese” structure cannot prevent the e5 pawn from queening.

Is there a moral to the story?

Actually yes: the two exchanges on d6 and e5 (both dark squares) are a thematic way of weakening
the opponent (here on dark squares) by diminishing his number of pawns on dark squares.

Is light/dark squares notion relevant in breakthrough, where the 'chess-like' closed pawn structures have no equivalent?

Yes, it is. In the example we just saw, Black had 7 pawns on light squares and 5 on dark squares.
After our thematic exchanges, 7 pawns on light squares and 3 on dark squares.
This is too unbalanced. And indeed, it's quite rare that a pawn on e5 is passed.

A pawn controls two squares (one, if placed on the edge) of the same color of its squares.
That basically means that if your balance is 7/3, you control only 6 dark squares
(and 14 light squares).

Rebalancing your structure takes time: at least two moves in the example 7/3, which
you can rarely afford.
And additional factor is that if you are unbalanced, your opponent usually “control”
the squares you are weak on.
Consider the toy example:
White: e5,…
Black: e6,f7,d7
Black to play cannot play on dark squares without being “en prise”.

One could argue or at least ask if this (equi-)partition of the squares (dark/light)
is not somewhat arbitrary?
Why don't we care about balancing the number of controlled squares on a,d,e,g files vs b,c,f,h files
for that matter?

The answer is simple: to defend we need to control consecutive squares (L1-distance beeing 1)
otherwise the opponent's pawns can breakthrough.
Thus, the balance dark/light is not arbitrary.

Let me give a further example of that theme.

[game;id:1599337;move:45;title:Black to play. Can you exploit the light square weaknesses? **]

A small quizz.
It is easy to see that one pawn can zugzwang three pawns but no more.
How many pawns can zugzwang two pawns?

Anyone?

If we just play against each other without exchanging ideas, I don't see the point.
Sometimes, we win. Sometimes, we lose. And what?