268 comments

[ 3.9 ms ] story [ 182 ms ] thread
I thought this would be about redesigning their movement to make a new kind of game. But it's only the shape of the pieces.

That said, I'd use a square for the rooks so they don't get confused with the bishops as they get moved and maybe an L shape for the knights. It would definitely make it easy for beginners to learn how each piece moves.

> It would definitely make it easy for beginners to learn how each piece moves.

Which is more or less the easiest part of the game to learn, and one which pretty much anybody will have down after one or maybe two games.

It's a clever little idea to share on Twitter but in practice I think it would make the board slower to read for advanced players, because so many pieces look similar.

Exactly, this will reduce readability of a board. It's like highlighting the black key notes on sheet music. You think you're making it easier, but end up just adding too much redundant info and it clutters the page. Or like adding a ton of crufty comments to code ("// this is a loop that ... "). Or like adding a "fill these. with water" post-it notes on an icecube tray. Clever. But worse.
If each movement is broken into a piecemeal function, then a non-standard piece could be assembled, such as a piece that moves like a knight in 3 directions and a bishop northwest and southwest.
This is fantastic, I would love to try playing a game with these pieces. I think this could improve accessibility and also make the game less intimidating for newcomers.
(comment deleted)
It would be interesting to do an experiment on people playing one another (online) with these pieces vs regular pieces (e.g. one player with a regular set and one player with this set). Because I do feel like I can visualize the state of the board more easily with this approach--but would that translate across people into better play, or is there some point where a more expert player (I don't play very often, though I enjoy playing and have played for years) has already internalized the mapping from piece shape to movement. But that said, it does feel like the difference between font styles in programming, which for me have a very meaningful impact over time.

Edit: Though good point to the parallel commenters, the knight shape is harder to differentiate and kind of throws me off. But maybe tweaks there.

The knight shape does seem off. But it avoids the piece being a sun symbol / fascist tribute piece :-)
If you're ever going to get the point where you recognize tactics (or anything, for that matter) on the board, how the pieces move must already be internalized and not require any thought. People even especially practice visualizing knight journeys between arbitrary squares to make this more automatic.

If you ask an experienced player to use a weird piece set, you will only be introducing error and cognitive overhead. If you ask someone naive to the game to use the piece set, do they internalize how the pieces move more quickly and get over the impedance when they start to use a normal set?

Going at a tangent to your idea, maybe it is possible to construct a set that aids the performance of a naive player, but which degrades the performance of a more experienced opponent. Could that handicap be as great as a club player offering pawn odds against a naive opponent?

Exactly this. Relatives have been trying to choose "fancy-looking" chess sets for me as a gift, and I always hated them. I don't want fancy pieces, I want the standard ones that don't give me cognitive overhead :-).

"Tournament chessboards" are my favourites, obviously.

Hell is filled with glass and ornate chess sets, I am convinced.
I've gotten so used to 2D knights that it annoys me when people face them forward on a 3D board.
There are flat chess pieces that have the traditional 2D symbols on top. Another question is whether anyone makes 3D knight pieces that are designed to lay flat on the board. Such pieces would fit a "Godfather" themed chess set rather well...
>Going at a tangent to your idea, maybe it is possible to construct a set that aids the performance of a naive player, but which degrades the performance of a more experienced opponent. Could that handicap be as great as a club player offering pawn odds against a naive opponent?

Blindfolded chess for the better player ;)

> has already internalized the mapping from piece shape to movement

Magnus Carlson, considered one of - if not the - best chess players of all time has a famous video of him playing three chess matches simultaneously, blindfolded, and winning all three.

Once you get to a certain point not only the movements but also entire gameplay strategies become internalized.

Not to mention the video where the board is presented in the "wrong" direction, so he enumerates to himself which squares the pieces are on, then closes his eyes so he can see the board clearer.
The more impressive one is where he's matched up against 10 Harvard Law students simultaneously, and is blindfolded. He won all of the games handily.

Even more impressive, one of the students asked Magnus if he could sign a chess board. Not only did Magnus sign an autograph, he literally annotated the entire game from memory. That is insane to me.

https://www.youtube.com/watch?v=w1Rr4Uq1R-I

Holy shit, I missed this one. That's wild.
Curious about the second tweet: are they typically called “liberties” in chess or is that borrowed from Go?
That is very clever!

But I hate the way the pawns face. Makes sense! But it still reads, to me, like one of those "color, but the text is in a different color" things. Feels like the pawns should be facing the opposite direction. =/

It's attack direction not movement direction. Pawns are the one piece where there is a difference and I'm not sure which is better ;-)
Also, there's rochade/castling.
Catapart is aware, they're just saying it's unfortunate how it ends up looking.
It is very clever, except accidental rotation can easily turn rooks into bishops and vice-versa.
Sounds like a fun variant, allowing that as a move... (Remicent of shogi).
You could put each piece on a square base that must remain aligned with the board.
Same. I understand why they're like that but I can't see it any other way than: "Why are pawns going to move in the wrong direction?".
I think change out \/ for:

    ■ ■
     P
And the knight for:

     ■ ■
    ■   ■
      N  
    ■   ■
     ■ ■
       
And you might have something a little nicer
Yeah, I wonder if it would be a slight improvement to combine move+capture.

So pawns would be

    /|\  /|\


    \|/  \|/

which looks less anti-directional.
It would make sense to have move and capture be different. Like square in front and two lines or points to the diagonal.
Also visualizes en-passant in a better way.
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.

"This is why rooks are worth 5 and bishops are worth 3"

Such evaluations are only a rule of thumb intended for beginners.

It's all context dependent. There are positions where a Bishop is more powerful than a Rook. Other things being equal, Bishops tend to be preferred over Knights in serious play (especially when you've got a Bishop pair, which is synergically more powerful than one Bishop times two), even though both are theoretically "worth" 3 pawns. Two minor pieces are usually worth more than a Rook and a pawn, and more often than not it isn't a good idea to exchange the former for the latter, even though the total value appears the same (3+3 = 5+1). And so on.

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.

At least at the level I play at, knights are invaluable because of their ability to jump over walls and threaten multiple pieces at the same time - they're the piece most likely to create a successful fork (threatening two pieces of value at the same time, likely forcing your opponent to pick which thing they're more willing to lose)
And it looks so much more reasonable, I would say
Except that this set describes the attack directions, not the motion directions. The difference is in the shape of the pawns. I think this one is quite interesting, as showing lines of threat.
The pawns felt like they were going backwards. Maybe if they had wee arrow heads.
(comment deleted)
Those peaces remain unreasonably expensive on Amazon, btw.
Peace has always come at a high cost. But if it's available on Amazon that still seems like progress.
Peace is dead; long live peace.
Won't fix the typo just so others could enjoy your joke
Good whittling project then.
(comment deleted)
I have the 1924 Bauhaus chess set and it's the most beautiful thing I own.

Always loved how the shape and size of the pieces indicate their moves while still being simple and recognisable.

If made into a physical set, I could see the bishops and rooks getting confused
Maybe it's just the top down view, but the white king and queen are practically indistinguishable to me.
I still have a set which was sold as "Visual Chess", and you can still find it on eBay for example. The pieces were mostly like traditional design (but kind of "modernized" in 1960s style) but when viewed from the top (i.e. as you play), there was an indication of the allowed move. You can see some examples in the image link below.

Even the pawns had a nice design which clearly indicated that movement of one square forward (normally), or attacking on the diagonal, were both allowed.

https://i.ebayimg.com/images/g/Yx4AAOSwTY9i7XeC/s-l960.jpg

https://i.ebayimg.com/images/g/AZUAAOSwuABi7XeE/s-l960.jpg

I had an updated version of this set (probably from the 90s?). Mine had a better board and I think the tops of the pieces were different, but the text/diagrams were identical:

https://www.reddit.com/r/mildlyinteresting/comments/c1ri78/t...

Not all of the moves were the clearest, but it was still handy while trying to learn.

That's the exact set I had. Worked pretty great for learning the game.
Great! This will be excellent for beginners. I'm considering teaching chess to my kids. Can I still find it online? My searches on Amazon and Google haven't yielded the results I need.
I'm pretty sure this was the set I had from 1992: https://www.ebay.com/itm/134677808785?hash=item1f5b6c4691:g:...

There's a lot of other versions sold under "chess teacher", this seems to be the most modern: https://www.ebay.com/itm/266236564204?hash=item3dfcef92ec:g:...

It appears that it was originally sold as "The Educator": https://www.ebay.com/itm/325899664936?hash=item4be1223e28:g:...

Not affiliated with any of those and they probably aren't the best prices (just the first ones I found), but should give you enough to track one down.

(comment deleted)
It doesn't represent the "en passant" rule, nor the castling. :-)
Nor the "promotion" of a pawn that has marched across the board to begin a new life as... a pawn (or anything else that isn't a king -- this almost means a queen, but a promotion to knight can also be useful).

Displaying usual the moves on the pieces is certainly super-handy for beginners to understand the moves, but as you've pointed out: Chess also has other nuances.

---

I was showing off one night and, at the tail end of an end game, and promoted a pawn to a pawn once it reached the far rank. It didn't really help in that game (although I can envision corner cases where it could be helpful as a blocker, but any other promotion is equally useful a blocker), but I had a move that I could afford to burn and it was fun to do so I did it.

It was a fun game. 10/10, did repeat other fun (and absurd) chess games.

> and promoted a pawn to a pawn

umm. Commendably warped, but I'm not sure that's legal?

(BTW many years ago I experimented creating chess variants, before I even knew it was a thing. I invented a number which all felt contrived, but one actually felt right and an improvement, which is the ability to take your own pieces, as well as the enemy's. Turns out that somebody's already invented it (search for 'self eliminator chess') and they considered it to have potential. The value of it was it allows you to trade off position for pieces. It really changed how the game felt for the better. FYI anyway, YMMV)

>> and promoted a pawn to a pawn

> umm. Commendably warped, but I'm not sure that's legal?

Interesting point. Here is what the FIDE handbook says about it (emphasis mine):

“When a pawn reaches the rank furthest from its starting position it must be exchanged as part of the same move on the same square for a new queen, rook, bishop or knight of the same colour. The player’s choice is not restricted to pieces that have been captured previously. This exchange of a pawn for another piece is called ‘promotion’ and the effect of the new piece is immediate.”

Source: Article 3.7.e of https://www.fide.com/FIDE/handbook/LawsOfChess.pdf

Any move is legal in casual play if the person on the other side of the board agrees that is is OK, especially if that game is held in their own house.

I mean: We're not playing ranked games here. It's supposed to be fun, and it was a fun game.

Self eliminator chess sounds like an interesting dynamic. I'll have to check it out. Some of the games where my thinking feels best (if that makes sense) are the games where a ravenous opponent has shewdly eaten most of my pieces.

Point taken, rules are as agreed between players, not divinely imposed. I was just curious about P-to-P promotion legality per the 'standard'

I hope SE chess lives up to its promise! It felt right and somehow freeing not to be always hemmed in by your own pieces.

I'm looking forward to absurd openings that include things like "I'm just going to use my bishop to get own pesky little pawn out of the way."

If chess is a war simulation, then: It seems like it is both brutal and useful to deliberately destroy one's own units.

(It's bad PR, always, but I'm both the PR department and the native public audience. It'll be fine on my side of the board, and it might even frighten the enemy.)

> this almost means a queen, but a promotion to knight can also be useful

So can a promotion to rook or a bishop, if for example promotion would cause a stalemate, especially if you're in a zugzwang.

Although such situations in actual real-world chess are _exceedingly_ rare, they make for a good puzzle material.

This idea is common in Shogi piece designs for those who can't read the traditional Kanji designs.
Indeed! Googling "shogi learning set" finds these sets.
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.
This would be awesome for teaching beginners.

I'd probably go with circle for the king, and maybe an L or a sideways H for the knight, even though they break the original design rule. Otherwise K/Q/N just look too similar.

+1 for making the knights an L. The way it is, the pawns sort of look like knights
Agreed. Love the concept, but the knights definitely need clarity.
Learning the movement is by far the easiest part of teaching beginners. It's fine as a design exercise, but its effect in making it easier to learn is close to null.
I have an overgrown dice pip renderer, so I made the rook "+", knight "L", bishop 4-gram, queen 8-gram, king octagon, and pawn a trapezoid (just to suggest movement direction). https://vezquex.github.io/clock/chess/
Off topic, but does anyone know a browser extension that lets you block certain domains from the HN list? I've blocked twitter/x from my DNS, and don't want links to it cluttering the feed.
I'm not aware of one, but if you can insert custom css rules it should be simple with something like

    tbody tr.athing:has(.title .titleline a[href*="twitter.com"]),
    tbody tr.athing:has(.title .titleline a[href*="twitter.com"]) + tr,
    tbody tr.athing:has(.title .titleline a[href\*="twitter.com"]) + tr + tr
    {
        display: none;
    }
This selects for any `tr` which has an anchor to a url with twitter.com somewhere in it (with some layers in between) and the two `tr` which follow. Note that this is pretty naive and could have some false positives with links such as https://example.com/blog/a-post-about-twitter.com and such, you could be more careful with the attribute selector if you'd like

This will leave "holes" in the numbering of the list of items as the numbers are not calculated based on the structure of the document but are rather hardcoded. You could definitely fix this with some more funky css rules if needed.

I would think you could automatically load the first match's "hide" URL until no more matches appear in the page, but it might trip some kind of rate limiter.
Pay/login-walled sites are common here.
(comment deleted)
Lichess also includes a piece set similar to this, although with the notable difference that the Knight is shaped like a greater-than symbol and the pawn is a small square.
Pawns are awkward though because movement doesn't align with attack
Yup, the visually dominating pieces end up _feeling_ wrong. It'd really bother me tbh.
a theme for chess.com would be nice !!

Lovely communicative design

I like it, but the knights still aren't quite right. And I'd probably give all of them square bases, so that the bishop and rooks weren't so easily mistakable with just a 45° turn.
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 the bishop look like it is "sniping" along the diagonals while the rook looks more "sweeping" maybe?

(c) don't make the king a little-queen. Make the king a little square to emphasize "it can only threaten the neighboring square," then it looks more visually distinctive.

The pawns are fun though.

Unrelated, but I just realized the horse is not in fact called "horse" like it is in Italian/Spanish.
When I was younger, I always believed that the proper name was for the knight was "horse" (my primary language is English). I only figured it out when I was playing chess against my friend, and he was like, "horse? wth is that"
It's been weird/amusing observing how different age groups at my local chess clubs call pieces and concepts. The older group very much stays to the local Dutch terminology for both while the younger group (presumably influenced by youtube) tends to mix Dutch piece names with English concepts. So you can have a koning (king) that is being "skewered" by a koningin (Queen) on the tweede (second) rank.
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.)

>an accidental misplacement turns a rook into a bishop or vice versa

This was my first thought too. If the "feet" of all of the pieces were "flat", or parallel to the rows of the board, you could more easily tell their proper orientation.

I also realized that if you defined the pieces like their moves, the knight's shape would be.... Interesting
I think I'd be inclined to disambiguate them by making the ends of the bishop pointed and the ends of the rook blunt. When moving along a diagonal you're moving in the direction of square corners, vs faces for moving in ranks and files, so it makes sense in that way, and it also evokes the shapes of the classical pieces.
A swastika shape would make sense for the knight.

I guess there is still a cultural aversion to its use, however.

A swastika?! Not quite, it's more like a cross: ꖅ
Speaking of swastikas, that's very close to an Iron Cross.

...anyway so people can see it better, that actual character is a pretty close representation of the knight's movement, it's just a little hard to tell the 2-then-1 distances: https://en.wiktionary.org/wiki/%EA%96%85

A swastika would only cover half of a knight's legal moves.
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 : )

I don't think (b) is really fair. You still call them "pieces" even if you are playing chess on a device/screen.
(comment deleted)