# Strange Material Balances?

I just wanted to look at some fun endgames, but all the popular ones (such as Queen vs. Rook or Minor Piece vs. Pawn or something like that) are already in thousands of textbooks! So I wanted some new material balances - that you guys get to make all by yourselves! So far, I've made up a few myself - such as two bishops and knight vs rook! So I guess I've said it all -- go make material balances in the comments!

Thats a material imbalance, and a win for black no matter who is to move.

True, (and i admit i should have written "imbalance") but if you were Black, could you make a win?

arceus2010 wrote:

True, (and i admit i should have written "imbalance") but if you were Black, could you make a win?

So basically you are asking us to put pictures of any possible positions on this thread?

Are you thinking of something like this? I believe it was a 90/30 game. There probably isn't much quality chess here... None of white or black was interested in a draw.

I've had queen & rook v. 2 rooks & knight come up a bunch of times lately.  The three pieces have yet to be beaten in this combo (I've been on both sides of it).

No i am looking for endgames with few pieces on the board that u think one side can win

Nowadays you can pop any position with 7-units or fewer into a tablebase and see what happens with perfect play by both sides. I tried a few settings where all the pieces are on their original squares, and the position below is the most interesting one I found. White has a forced mate in 90 moves.

Of course White has multiple ways to win at each step, but if we treat this as a directmate problem where White must find the quickest way to mate, a surprisingly large number of white moves are uniquely forced. In this main variation where I tried to maximise the number of such forced moves, 77 of the 90 white moves are unique and they are given with an '!' in the analysis. Since only 13 white moves have alternatives that work equally well, these dual moves are mentioned as well whenever they appear.

As in most tablebase results with long play, the solution is not really human-comprehensible. But out of all these best possible moves, the stand-outs are three instances where the WK retreats from e2 to d1 for no apparent reasons. If this were a composed problem, these recurring Kd1! would have been an amazing theme!

Rocky64, exactly post stuff like that! Also, do you know any good tablebases? The one you used, by the way, was Queen and Knight vs. Bishop, Knight, and Rook.

You could also put two engines up against each other in the position and see how they play.

Wat is sad is the 50 move rule

The Lomonosov tablebases cover all possible positions with 7 units or fewer and so is the best. AFAIK, the only way to access it for free is through an Android app. Check their site for details: http://tb7.chessok.com/

For positions with 6 units or fewer, you can access the Nalimov tablebases for free here: http://www.k4it.de/index.php?lang=en&topic=egtb

It looked like a total crush for White, but for some reason, most of the moves were draws!!! Any ideas why?!!

That's a position where White has mate-in-67, but start the R on b1 instead, say, and it's a draw. So yes, it's bizarre but typically so for tablebase results and not really possible to understand. Still, you can play through some of the drawing lines and get some ideas of how Black does it.

W 5points/B 9points

I luckily found an online tablebase!!

https://syzygy-tables.info/

I once ended up with something like this against my computer.  It was no fun, believe me...

ghost_of_pushwood wrote:

I once ended up with something like this against my computer.  It was no fun, believe me...

That's a draw according to the tablebases, but some WK-moves allow Black to win. The most interesting case occurs after 1.Kh2, when Black has only one precise way to win and hence it works as a puzzle. The sequence is quite short too and understandable! So, how can Black win the WQ?

As I said, the position was something like that...   And I am not tablebases (thus, it was not a draw).  I lasted maybe 25-30 moves.