Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I realized chess pieces can be redesigned to be geometric attack directions

twitter.com

Post reply on HN
Print view View original

221–230 of 269 posts

Re: I realized chess pieces can be redesigned to be geometric attack directions

#221

Post by V-2 »

V-2 wrote: post #193

Fun as it may be as a thought/UX experiment, I don't think it's a good design, because half-truth is usually the worst compromise. This design doesn't account for many special qualities and behaviors pieces exhibit. It doesn't indicate that * while pawns do capture diagonally (forward), they are pushed forward in straight line * they can capture en passant * they can move by 2 squares from their initial file, and onl…

> * they can move by 2 squares from their initial file, and only from the initial file

Obviously I meant a rank (horizontal), not a file (vertical), sorry. Cannot edit the comment anymore

Re: I realized chess pieces can be redesigned to be geometric attack directions

#222

Post by NeoTar »

tauchunfall wrote: post #192
cruano wrote: post #58

Earlier quoted context omitted.

Unrelated, but I just realized the horse is not in fact called "horse" like it is in Italian/Spanish.

It is called "Springer" (translates to: jumper) in German. Sometimes "Pferd" or "Ross" is also used (translates to: horse and steed, resp.). That's why "Knight's tour" is called "Springerproblem" or "Rösselsprung" (from "Ross") in German.

Hence the German academic publisher, Springer, which uses a knight as it's logo.

(Not to be confused with the much larger Axel-Springer publishers, nor the American academic Springer publisher.)

Re: I realized chess pieces can be redesigned to be geometric attack directions

#223

Post by V-2 »

andoando wrote: post #95
infogulch wrote: post #63

For me the most surprising observation from this thread is this tweet [1] showing that the knight's moves are exactly the squares that the queen can't get to within a 2 square manhattan distance. I.e. the knight is an anti-queen. [1]: https://twitter.com/skidbladnirr_/status/1750285122769957129

To add to that, this also makes a knight and queen a very strong attacking combo, because they can uniquely cover the most squares.

It also enables some powerful, well-known attacking patterns, like a smothered checkmate with Queen sacrifice (or Philidor's mate).

Re: I realized chess pieces can be redesigned to be geometric attack directions

#224

Post by V-2 »

dudul wrote: post #151
infogulch wrote: post #63

For me the most surprising observation from this thread is this tweet [1] showing that the knight's moves are exactly the squares that the queen can't get to within a 2 square manhattan distance. I.e. the knight is an anti-queen. [1]: https://twitter.com/skidbladnirr_/status/1750285122769957129

I'm not a master or anything, but I've always found it interesting that the queen is usually thought of as the most powerful piece while I think it's the knight. Yeah the queen can move everywhere, but she's "predictable". The way a knight moves makes it more difficult to defend against IMO.

Knights tend to be overrated at lower levels of playing strength, where their irregular moving pattern is indeed challenging for the players, and as a result, Knight moves often get overlooked.

However, once a player becomes more proficient, not only the obvious superiority of the Queen, but even of the Bishop (other things being equal, Bishops are thought to be actually somewhat more valuable, especially as a pair) starts becoming apparent.

Re: I realized chess pieces can be redesigned to be geometric attack directions

#225

Post by V-2 »

dev1ycan wrote: post #145

What I find funny about chess it's such an elegantly designed game... except castling. Always seemed like a random thing to me to be in the game, and then I found out that it was added centuries after in medieval europe to the game, makes sense why it feels so out of place. Wish we had a league without castling for that sole purpose.

No-castling variant is a thing. It generally makes taking the King to safety more difficult, encouraging a more aggressive style (in the spirit of "I'll get to yours before you get to mine").

Kramnik revived the idea during his collaboration with DeepMind (where he tested several chess variants, revolving around only minor, albeit impactful adjustments to the ruleset).

One potential problem with no castling is that players may end up castling "manually", which takes several moves and becomes a mundane "housekeeping" task, actually sucking some dynamics out of the game. That was probably the reason why it was invented in the first place, to sort of automate the obvious thing away.

By the way, European chess added much more to the original version beyond just castling. The Queen would originally move in the same manner as the King (only by one square), the "mad Queen" thing is a European invention. So is the pawns' ability to move forward by 2 squares from their initial position. (And I'm guessing the incentive for that was of similar nature as above - to speed the game up).

Re: I realized chess pieces can be redesigned to be geometric attack directions

#226

Post by mtlmtlmtlmtl »

boredhedgehog wrote: post #204
LeegleechN wrote: post #26

This idea is common in Shogi piece designs for those who can't read the traditional Kanji designs.

And the idea is misguided there as well. Kanji are problematic because people who can't read them can't differentiate their shapes. Any unique silhouette -- even a Latin letter -- is preferable to arrow diagrams, which end up unnecessarily similar to one another.

Yeah I had no issue learning Shogi with no knowledge of kanji.

Re: I realized chess pieces can be redesigned to be geometric attack directions

#227

Post by V-2 »

Reason077 wrote: post #120
crdrost wrote: post #52

I think this is fun, but (a) knights' threats are disconnected and this "flower" approach while good is maybe not ideal. You might as well use a hollow circle at that point. (b) when we say "chess pieces can be redesigned to be..." then I think about actual physical pieces, it would not do to make these as actual physical pieces because an accidental misplacement turns a rook into a bishop or vice versa. Gotta make t…

A swastika shape would make sense for the knight. I guess there is still a cultural aversion to its use, however.

It would need 8 arms, not 4. (A well-placed Knight meaningfully controlling all of its 8 target squares is actually referred to as an "octopus Knight" sometimes).

It is also opinionated, in the sense that the vectors for Knight moves is [2, 1] and [1, 2] (whereby in both cases negative numbers can be used)... so whether you visualize a Knight move towards the left top corner as an L (x-1 first, then y-2), or as an ꓶ (y-2 first, then x-1) is an arbitrary decision.

It might be my OCD side coming out, but any design prioritizing one of these perspectives over the equally valid alternative doesn't feel "clean" to me somehow : )

Re: I realized chess pieces can be redesigned to be geometric attack directions

#228

Post by fasteo »

infogulch wrote: post #63

For me the most surprising observation from this thread is this tweet [1] showing that the knight's moves are exactly the squares that the queen can't get to within a 2 square manhattan distance. I.e. the knight is an anti-queen. [1]: https://twitter.com/skidbladnirr_/status/1750285122769957129

>>> I.e. the knight is an anti-queen.

The complement-queen I would say

Re: I realized chess pieces can be redesigned to be geometric attack directions

#230

Post by mtlmtlmtlmtl »

neilk wrote: post #14
mxwsn wrote: post #10

What I enjoy about this design is it exploits the correlation between piece value with the number of ways it can move, resulting in a visual intuition that more visually complex pieces are worth more on average. The knight is perhaps the frustrating exceptions to this.

Piece value is determined by how many squares it can threaten. This is why rooks are worth 5 and bishops are worth 3, despite both having apparently equal freedom of movement (just rotated 45 degrees). The bishop will be able to access strictly fewer squares than the rook even in the best case, and usually much fewer. And the bishop is stuck on one color.

Actually the numbers 5 and 3 are not derived from the number of squares attacked.

They denote the power of pieces in relation to eachother, as a general rule with lots of exceptions, and reflects many material imbalances that are often comparable in strength(usually in the endgame). The unit of measurement here is the material value of a single pawn.

3(3x1) pawns are often enough compensation for for a knight(3) or a bishop(3 + ε).

3 minor pieces(3×3) are often sufficient compensation for a queen(9).

One minor piece and 2 pawns(3+2) are often sufficient for a rook(5).

Two rooks(2x5) are typically slightly stronger than a queen(9).

Two minor pieces(2x3) are typically slightly stronger than a rook(5).

There are different systems that are derived from number of squares attacked. One thing to note about these systems is that pawns are not created equal. Rook pawns attack one squares less than other pawns, for instance.

Post reply on HN