Minor Pieces vs Queen

@61

The king is unhappy in the corner.
It would be interesting to query if such a material has ever occured in a real game.

There are also positions with the king off the edge. The queen draws here, for example

And here

I imagine the frequency of the endgame in real games is rather small, but I would guess it has occurred at least thousands of times. (Many games have been played since chess was invented.)

@64

"I would guess it has occurred at least thousands of times"
++ I doubt it. The queen can be a promoted pawn, but it is rare for BBNN to survive trades.

I think there are about 3000 possible 7 man endgames, obviously not all occurring with the same frequency. If this occurs with 1/10,000 the average frequency of an endgame then you could expect around 1 in 30,000,000 games that reach a 7 man endgame to reach this one.

I'd guess there have been well over 30,000,000,000 games of chess played that have reached a 7 man endgame since it's invention.

A majority are probably played by beginners when the remaining material is something of a lottery, so the assumption of a frequency of occurrence of 1/10,000 the average is probably very conservative.

At any rate I'd say the chance it has never occurred is as near nil as makes no difference.

I think these endings are too hard to succumb to a simple materialistic approach. If it becomes 3 pieces v. queen the queen doesn't necessarily save the game because around 20% of positions are still winning for the 3 pieces. You have to be capable of evaluating the 3 pieces vs. queen positions to play it accurately and for that you need similarly to evaluate the possible 2 pieces vs. queen positions that may result. But in practical play even the 2 pieces vs. queen positions usually prove too hard. According to tablebases the great majority of such positions are won for the queen, but in practice the usual result is a draw.

This position, for example, is a win for the queen under basic rules. A player could be tempted (as Stockfish is) to have two of Black's pieces away immediately.

But that blunders half a point as soon as he takes the first knight.

A human player who has looked at 2 minor pieces vs. queen endings may well avoid that trap but I strongly suspect that almost all players would fail to bring in the win against a tablebase.

You could say Stockfish can be excused because it's designed to play competition rules and the position is already drawn under competition rules, but it does no better from this position which is a White win under both basic and competition rules.

You have previously said SF is weak on these endgames, but I would invite you to try playing White against SF from the same position and see if you can do any better than 5-5.

(I think this also demonstrates how useless the theoretical figures I've posted are when applied to practical play.)

Sometimes 3 minor pieces > queen in the endgame.

@AmericanUnderdog

About 20% of positions, as I said in the previous post. And vice versa about 50%. But actually winning with the queen probably won't happen 50% of the time. Try winning with the queen against the Syzygy tablebase from this winning position.

Or just against an engine via the magnifying glass.

If you fail against the engine, don't be downhearted. So does the engine.

Thank you for elaborating on this.

Out of interest here's an example of how the above position should be played (Rybka has the Nalimov tables).

(But don't ask me to annotate it.)

Hi, guys

The 2nd position is a fortress, Notice how the bishop and knight protect the borders of their king's 3 by 3 box in the corner. King can not penetrate throught that. so draw it is

The 4 standard minor pieces, as well as 4 knights, both have been proven to beat a queen.

Queen wins 3 minors more often than it loses. And mathematically.

Rook, Knight, Bishop beats Queen.

This conversation is so close to sounding incredibly illegal.

Maybe Rook + 2 opposite colored bishops, but I think RBN is a draw.

If you look further up the page I posted three positions with 4 minor pieces v. queen that are theoretical draws and two that are won by the queen, so however it was proven, it was proven wrong.

It is, however, easy to rigorously prove that there are no combinations of material that contain any legal positions and are theoretically won for either side from all positions.

KQKRBN is drawn from 47.2% of positions with ply count 0 under the 50 move rule under either FIDE basic or FIDE competition rules.

Yeah I know I always mean from the fairest starting positions without any immediate tactics.

These are the types of positions I test for various piece combinations, to make it start off as balanced as possible:

Even if it looks like this?

yh