What is the most complicated endgame

Sort:
Avatar of MadMuppets
n9531l wrote:
eveningstarandlion wrote:
However, I do suggest you try Queen Knight vs Rook Bishop Knight. You have NO idea how long it takes to win. Like, 200 moves maybe.

I decided to set up some random Q+N vs. R+B+N (White to move) positions to see how they came out. One of the longer mates (ignoring the 50 move rule) is shown below. In this position, if the white knight starts on c2 instead of b1, the position is a draw. I don't know what the longest mate is.

White to play and mate in 388.

It is actually mate in 289, but still impossible to understand and train the technique. I tried to understand 2 Knights vs pawn and Queen vs Rook (just for fun, because I don't think it is that useful). These endgames are much more difficult than most people think, but are trainable.

Avatar of MARattigan
SmyslovFan wrote:

Take a look at the following endgame played last week. Try working it all out without the use of an engine. It's insanely complicated!

And an engine won't be much help if you want the right answer. SF can't even do KNNvKP.

Avatar of MARattigan
MadMuppets wrote:
n9531l wrote:
eveningstarandlion wrote:
However, I do suggest you try Queen Knight vs Rook Bishop Knight. You have NO idea how long it takes to win. Like, 200 moves maybe.

I decided to set up some random Q+N vs. R+B+N (White to move) positions to see how they came out. One of the longer mates (ignoring the 50 move rule) is shown below. In this position, if the white knight starts on c2 instead of b1, the position is a draw. I don't know what the longest mate is.

White to play and mate in 388.

It is actually mate in 289, but still impossible to understand and train the technique. I tried to understand 2 Knights vs pawn and Queen vs Rook (just for fun, because I don't think it is that useful). These endgames are much more difficult than most people think, but are trainable.

No it's not mate in 289.

If the game is played under FIDE competition rules it's a draw. See here. The Lomonosov tables are no longer with us, but the link given shows that it can't be a win in less than 359 moves under FIDE basic rules either.

KNNvKP is trainable under FIDE basic rules, but I very much doubt under FIDE competition rules. I would happily wager my house and spouse that not a single person or engine in the world could reliably win all winning positions in the endgame against Syzygy under competition rules (not too risky since I don't have either).

Avatar of Alexandrehanher

😊

Avatar of idk13243

King two knights vs king

Avatar of MARattigan

Well it's usually difficult to win for either side, but you really couldn't call it complicated.

Avatar of Mariajoffil

KQ vs KR

Avatar of elz50

KNN vs KP

Avatar of Lord_Darth_Vader_7

Queen vs two rooks is kinda complicated endgame.