I had the last week, a very extensive look to the famous pawnendgame using many sources.
- Megadatabase games and analysis
- Correspondence games and analysis
- Analysis of chesspub
- Analysis of Argentinian Luis Ernesto Rodi Maletich:
http://www.ajedrez.com.ar/rodi_7.htm- ...
My computerengines were quite helpless in this endgame as often the variations were leading far beyond the horizon of the calculations.
The endgame is extremely complicated and even after many hours analysis and using all these resources, I don't dare to say that my analysis are bulletproof. I found however plenty of ameliorations on old analysis and many new ideas and tricks.
I believe white has the following mainideas for a win:
- Putting king on a6
- Putting pawn on f6 and then breaking through with the king to pick up the pawn on f7
- Create a free pawn on the queenside and win via triangulations.
Blacks defences are in choosing the right setup on the queenside:
- Blocking with a5 even if this gives a passed b5 pawn for white
- Giving white the free pawn on the queenside but generating sufficient counterplay with d5
- Keep the a-pawn on a7 and organise the defense from the 7the line.
Key I think for both sides, is to keep maximum flexibility. On the other hand if white wants to win, he will need to move forward.
I tried to use a very recent mastergame as tool to explain all the ideas.
White became GM next year. Black became IM next year and won on top Moscow Open:
http://www.chessvibes.com/reports/untitled-vladimir-belous-wins-moscow-open/Chess is a very complex game if even such simple looking positions can lead to so many mistakes.