3

I recently found a chess problem YouTube video with this position.

[FEN "r4rq1/5p1p/p5kn/R1B1Q1b1/6p1/5p2/PNP2P2/1K6 w - - 0 1"]

I want to turn this position into an entire game of chess. Does anybody know of a way that I can generate moves that would lead up to that position to make a full game? I don't want to find out the exact moves that were played if this position ever happened. However, I want to find out if I could make this into an entire chess game that looks legitimate.

Rewan Demontay
  • 17,514
  • 4
  • 67
  • 113
SteepAtticStairs
  • 39
  • 1

2 Answers2

3

It is very easy to generate a "proof game" for this positions (and if you are too lazy, you can use problemists tools like Natch or Jacobi, but I think it will be very hard to make the proof game look "natural". (It gets worse as the position itself is rather unnatural. Did the Kg6 have a death wish? Maybe if the last move pair was Rf6xg6+!, it looks a bit more believable, but still the Qg8 is ridiculous.)

Isac
  • 703
  • 7
  • 16
Hauke Reddmann
  • 16,242
  • 3
  • 27
  • 70
2

I think the best approach for loose positions like this without retro interest is to have three structured phases:

  1. Knights marraud to eliminate captured pieces not required for pawn balance. If necessary, c & g pawns edge forward to avoid check being given. This gives a homebase diagram.
  2. Move pawns to final positions, including captures. This gives diagram A.
  3. If it's not obvious, use Jacobi in demolition mode to go from A to B (here in something like 10.0 moves).

This approach can obviously be generalized. But if it fails to work, it will fail to work for a specific reason, and you can see whether this is a fundamental blocker or the solution approach has to be adapted.

Laska
  • 11,846
  • 4
  • 44
  • 76