Chess Chats

This is my template to record chess position.

En passant square and unmoved King/Rook are marked by dots as illustrated.

1 Like

Black to Mate in 6

1 Like

I’ve started studying Strategy/Positional. Looks like I need a white-out pen. :sweat_smile: I like this column format better. Still missing out on indented combo notation, though.

The difference between tactical and positional, combination and strategy.

Mate in 33? I’d probably claim a mate in 35 just to give me a leeway. The point is, I’ll sacrifice all my pieces doing it!

1 Like

Fresh off reading Aron Nimzowitsch’s My System and I play this game. Notice how the moves I did anticipated Black’s moves!

Also a lesson in space management vs premature queen. (vs William Stewart’s Chess Psychology: The Will to Win! )

I get better with my concentration. I win 1 of 4 matches against a 1300 bot on chess.com. Fun!

1 Like

I’m trying out Chess Informant notation style. It’s certainly compact, but I’m not sure I should adopt it. Maybe I should use the full algebraic notation, instead? I used to do that.

Looks good. And the short notation is fine. I don’t know what Chess Informant notation means. Maybe using a symbol instead of a letter? If so, then I think it’s good.

It omits the marks for checks and captures. Here’s more info on it.

1 Like

Yes, beating and checking don’t need to be written down. They are implicit. The only extra notation that is sometimes neccessary is when two pieces of the same kind can reach the same field.

And about me playing against ELO 1300 bot: It nearly never makes an “idiot move” (like leaving the queen to be beaten) at this level.

1 Like

That’s right. It’s still vulnerable to a combo punch, though. Anyway, I manage to “trap” the Queen on this game. Seems to be vogue to lose the Queen these days. I even caught Anna Cramling lost hers during a Twitch live stream!

1 Like

I got new markers with bolder cool colors. Just testing them out. Look at all those checkmates I’m missing. :slightly_frowning_face:

1 Like

Mate in 37!? Sounds impressive until you realize the computer is wrong! You can mate faster than that! Say, in about 20 moves? The end game database says it’s mate in 15.

Here’s the game. It has 6 great moves, and what’s that King doing ahead of every other pieces? :rofl:

Ronfuse vs rangstorm -

1 Like

The computer is not wrong, it is just saying it proved a mate in 37, not excluding faster ones. Depending on settings, many chess engines stop searching as soon as they find a mate.

I personally would consider that wrong.

But it’s nice to find some moves that are more efficient than the AI. Though that only happens because the AI is made artificially dumb.

I would always loose if the AI would run “full power”.

1 Like

The engine reports the mate when it exhausts all possible variations for one move. To find a faster mate it will have to exhaust many other moves, something potentially time consuming so in many cases it stops early (or maybe the allocated time runs out) even when playing with full strength. There are different modes used when analyzing positions where the search will continue, and usually also specific modes for mate search, where of course stopping would indeed be wrong (but will still likely find a few longer mates first).

1 Like

Well, for one thing, you can skip directly to move 56. If the computer didn’t even check for that transposition, that’s some lazy coding, there.

In any case, I don’t think Analysis mode is handicapped in anyway except depth. I think 18 ply in this case. Therefore, it makes sense that the computer should analyze all checks and captures and find the shorter path.

From the way you describe it, it sounds more like depth first search instead of breadth first search.

The depth in modern chess engines is largely an approximation, as there are several heuristics and extensions run on top of the normal search (and the search is not exactly breadth fist, but this is a huge topic). In my reply above I was assuming a time limited search as I don’t know what chess.com is doing for analysis - obviously with infinite analysis the best mate would eventually be found. I checked on lichess and it found a mate in 20 first, then almost immediately one in 15 and rather quickly the final one in 10.

1 Like

Let’s say a chess variant where you play as one of the pieces. You lose if your chosen piece gets captured, or the King gets checkmated. Your choice is kept secret from your opponent, so determining which piece is the chosen one is done while playing the game.

Which piece would you choose? Noobs would choose to be the Queen, as did I in the beginning. Nowadays, I’d choose g pawn as the most likely to survive the Endgame. Else, the a rook.

Here’s a sample game:

Black chose the Queen. White chose the g2 pawn.
It’s interesting to see how the game changes when you selfishly try to keep your chosen piece alive!

Somehow the computer grossly underestimate my attack. It’s at least winning the Queen if not outright mating combo!

Full game here: