What's the idea of 'all rook endgames are drawn' ? i think opposite coloured bishop endgames and even queen endgames are more drawish

Question

Quote: 'all rook endgames are drawn', 'All rook endings are drawn' or 'All rook and pawn endings are drawn.'

• For queen endgames, Karsten Müller even gives a way of visualising this (not quite Karsten Müller's point or intention, but at least the ff is my interpretation): Imagine some arbitrary queen endgame or rook endgame and then imagine the queens are replaced with rooks or vice-versa. 1 example was for a position involving queen and 2 pawns vs queen. Karsten Müller points out that the position was drawn but then it's not drawn for if the queens were replaced with rooks.

• For opposite coloured bishop endgames (eg this recent question), Josh Waitzkin gives examples how they are so drawish. Even in my own games, this is the easiest kind of endgame to see its (insane) drawish tendency. I absolutely do not see queen endgames or rook endgames on the same level as opposite coloured bishop endgames.

Question: What's the idea of 'all rook endgames are drawn' ?

My guess:

I suppose even though opposite coloured bishop endgames are like superGM level in terms of drawish tendencies doesn't mean rook endgames can't be IM level or even GM level (like rook endgames are 90% drawish vs queen endgames or opposite coloured bishop endgames are 99% drawish?).

Perhaps the idea of the quote is that it's difficult to convert a 1 pawn advantage (or that 1 pawn isn't much of an advantage. or is it?) in rook endgames, but then the difficulty here doesn't mean there aren't more difficult endgames to win (eg 2 pawns up in either queen endgame or opposite coloured bishop endgame).

Attempt of answering the question by statistics:

Are there any statistics or something to show how drawish rook endgames are? Otherwise I guess the answer to this post is gonna be just

The context of this quote shows it is a comment on the fact that a small advantage in a rook and pawn endgame is less likely to be converted into a win.

Answer 1

I try my best:

Declare a pure pawn endgame as comparison basis.

• A single plus pawn may be enough to win.
• Two plus pawns win regularly (unless it's a very unlucky situation).
• A large positional advantage may also win.
• The chances generally rise with more pawns. (You can send the enemy king after your passer while gobbling elsewhere.

Add two knights.

• "Knight endgames are pawn endgames." (Botwinnik)
• The above statements still hold approximately.

Add two like bishops.

• Ditto, as you can trade them.

Add two unlike bishops.

• Extremely drawish because of blockade.

Add two queens.

• Extremely drawish because of the weaker side giving a hail of checks on your kings crown.

Add two rooks.

• Yes, the jocular saying goes that all rook endgames are drawn (and statistics is in favor of that), but unlike queen and different-colored bishop, it is hard to pinpoint it to an one-dimensional reason. Be careful what you compare - Ke3 Pe4/Ke6 is drawn whether alone, +N,+B,+R or +Qs. Ke3 Pe4/Ka1 is (usually) still won only with +R.

I daresay rook endings are more difficult than the other kinds.