Comments/Ratings for a Single Item

here's a quick first draft to see what it would look like: select 2d motif in the drop down.

Right now, you're using the pieces that are upsidedown for the first player. I would recommend including both sets, so that the player has the option of choosing which set is rightsideup for the side he is playing.

transparency need to be improved.

I had meant to make them transparent, but I suppose I forgot a step in the process. So I reloaded the images, made sure to complete all steps, and tested the results for transparency. You just need to refresh your browser cache to see them.

So in short: doing what you propose would be approximately as difficult for me as it would be for you to do it yourself, and comparable to the task of rephrasing one sentence in the Chinese translation of the bible...

The simpler solution is to provide the set in two different orientations and let the player choose which one is rightsideup for him.

That's a good idea. The 'Classical' representation also suffers from the flipping problem, and this could be used there too.

Note that even the Japanese don't draw tiles around the kanji when they publish diagrams (e.g. tsume puzzles).

Traditionally, these were published in print media, such as newspapers and books. Since they were already printing Japanese text, it wouldn't be difficult for them to print half-Kanji piece names, whereas it would probably take extra equipment, such as specially made typefaces, to print full representations of Shogi pieces. It would probably also take up more space if they printed full representations of pieces in their varying sizes, and in print media, extra space costs more. At least, Japanese readers could easily tell which pieces belonged to each side by their orientation. For them, this would be as easy as it is for us to tell when Latin characters are upsidedown, and maybe even easier, as we have some letters that look the same upsidedown (O, I, H, X) and some that look like each other upsidedown (M and W, p and d). For westerners with less familiarity with the Japanese language, it helps to have pieces in wedge shapes. A computer interface for playing Shogi does not have the limitations of print media, it is not targeted at just a Japanese audience, and since it is used for playing the game instead of just solving puzzles, every way it can simulate the experience of playing across the board helps.

Well, it would be very easy for westerners and Japanese alike to see whether the traditional glyphs for chess pieces are printed upside down, so there is just as little need for drawing tiles around them as it is for the Japanes to draw them around ther kanji. And space is more cramped on a computer (or telephone!) display than on a newspaper page, so that would be another important reason for not drawing them.

Simulating the experience of playing across the board helps??? To chase people away, most likely! Playing Shogi across the board is an absolutely horrible experience. It is the main reason why the Dutch Chess Association has about 150,000 members, and the Dutch Shogi Association only 56...

Well, it would be very easy for westerners and Japanese alike to see whether the traditional glyphs for chess pieces are printed upside down

Only for westerners with enough experience using Japanese pieces. For beginners or for players who choose to stick with western pieces, not so much.

And space is more cramped on a computer (or telephone!) display than on a newspaper page

Yes, space is cramped on a phone. Even when I loaded Jocly into my iPad earlier today, the display of the board and pieces was too small. However, the issue I raised was not that newspaper pages were cramped. It was that everything printed on a newspaper costs money to print. My desktop monitor, while still smaller than a newspaper page, is more than large enough to comfortably play Shogi.

Simulating the experience of playing across the board helps???

I should say simulate or enhance with a preference for enhance. Computers can enhance the experience of playing Shogi by enforcing rules, showing legal moves, and letting each player use a set he is comfortable with. But using evenly-sized half-Kanji characters without wedge shapes is not an enhancement unless that is how a particular player prefers to play, which might help on a phone, though it's harder to imagine anyone would have a preference for them on a larger display.

Only for westerners with enough experience using Japanese pieces.

I was not talking about japanese pieces. The issue is whether people without any experience whatsoever could see the difference between this  and this , or whether they would need help of a surrounding pentangle before they could spot the difference. Seems to me the pentagle is totally useless. Of course it would be far easier still to see the difference between this  and this .

There is no accounting for tastes, but some representations are just objectively inferior. This could be measured by how long it takes people not familiar with any of the representations under test can see the solution to trivial problems (like "what is the best capture in this position", while PxQ is possible). That there are people that like the inferior doesn't make it any less inferior.

Seems to me the pentagle is totally useless.

In my Symbolic set, it's the only thing that will distinguish between Pawns or Gold Generals on each side.

While they could more easily tell Silver Generals apart, it would take some learning, but the wedge shapes help a player immediately tell which is which.

If it weren't for the baselines in the Motif set, it would have similar problems. But even when players can recognize the orientation of each piece without wedge shapes, these shapes provide a consistent means of recognizing which side is which that doesn't require further knowledge of the individual pieces, and that can be helpful even when you do already recognize the pieces.

Also, in the western set I criticized, the Silver and Gold General images would be too unfamiliar to players for them to easily tell which is on which side right away if only orientation was used to distinguish them. When I first looked at them, it seemed like they were oriented differently than the other pieces, and it took me a while to figure out that they probably represented epaulettes on a shoulder.

In my Symbolic set, it's the only thing that will distinguish between Pawns or Gold Generals on each side.

Well, that just shows how unsuitable the set is. You would not have that problem if one of the two players used black pieces.

Also, in the western set I criticized, the Silver and Gold General images would be too unfamiliar to players for them to easily tell which is on which side right away if only orientation was used to distinguish them. When I first looked at them, it seemed like they were oriented differently than the other pieces, and it took me a while to figure out that they probably represented epaulettes on a shoulder.

The point is of course that orientation is not used to distinguish them at all. They will never be displayed upside-down, but one is white and the other black. The stars are just to distinguish the 'rank' of the generals; the more stars the higher. And this number happens to coincide with the number of non-forward moves, which offers the oppotunity to place them in a pattern that is slightly mnemonic for the direction of these moves in the white POV.

Well, that just shows how unsuitable the set is. You would not have that problem if one of the two players used black pieces.

No, it is not unsuitable at all, and it is not a problem, for it follows the Shogi convention of using pointed wedges to distinguish sides instead of the Chess convention of coloring pieces differently.

Yes, and that is exactly what makes over-the-board Shogi such a terrible game. That convention is highly inferior.

Yes, and that is exactly what makes over-the-board Shogi such a terrible game. That convention is highly inferior.

Well, this helps illuminate what you meant when you said this:

Playing Shogi across the board is an absolutely horrible experience.

Anyway, I disagree with you. I find the Shogi convention better for Shogi and the Chess convention better for Chess. At least with computers we can let players use whichever convention they prefer.

Anyway, I disagree with you. I find the Shogi convention better for Shogi and the Chess convention better for Chess. At least with computers we can let players use whichever convention they prefer.

Well, as I pointed out, it is not a matter of opinion. What is a better representation for playing a game is a verifyable fact, measurable by the average rating drop or raise people would get when having to play with one representation or another. E.g. representing all pieces by empty squares (aka blindfold chess) usually greatly degrades the rating of non-GM players, and can thus be considered an unsuitable representation for playing Chess.

And 'disagreeing with facts' is also known as 'being wrong'.

For people that want to experience the difference between using pictogram and tile pieces, I set up a simple speed test cq dexterity game here.

The idea that Shogi and Chess would have different needs does seem rather inconsistent, as they are practically the same game. I have difficulty identifying any aspect of either game that would cause such a difference.

But it is indeed good that people can choose their representation on a computer independently of that used by the opponent. If there is at least one suitable representation amongst the choices.

Dear friends, I'm following your interesting discussion. From my perspective, I feel that you are both right with your respective arguments, difficult to say.

HG, I found your little game quite funny. Well, I did 2.6 for pictograms and 4.3 for kanjis, but is not so meaningful. First, I am an European, not used to see kanjis. I guess a Japanese or a Chinese would go faster. Second, it just mean that color is perceived quicker than the orientation of a wedged shape. So what? Chess nor chu shogi is a game of speed, it doesn't matter if my brain needs 2s more to identify a situation. Third, the problem just focuses on identifying an anomaly in the setup, things can be quite different for a full evaluation of the situation on a board. There is difference between the two systems might be reduced to an unsignificant number. I mean if I need 1 mn to evaluate a full board, spending few seconds more because of the color-or-orientation system is not a problem per se. Especially if I were an Eastern-educated person.

Now, speaking of my personal experience, I just played once shogi on-board, against my French publisher, who was more at ease than me with kanjis, and it was a ... massacre. I remember being surprised of how difficult it was to identify the opponent pieces because the kanjis were upside-down for me. Even though I knew of all them. I concluded that I would probably need a 1-kanji set instead of a 2-kanji set. And, I confess, I bought a second set but with ... pictograms! Shame on me.

Well, as I pointed out, it is not a matter of opinion. What is a better representation for playing a game is a verifyable fact, measurable by the average rating drop or raise people would get when having to play with one representation or another. E.g. representing all pieces by empty squares (aka blindfold chess) usually greatly degrades the rating of non-GM players, and can thus be considered an unsuitable representation for playing Chess.

And 'disagreeing with facts' is also known as 'being wrong'.

The pictograms are better at distinguishing the sides visually (e.g. the Mnemonic pieces for the large variants). However, Shogi uses Kanji pieces by tradition, and it has been this way even before the drop rule was added. Most Japanese players use this system, and even Western players (such as myself) have a tendency to use the kanji system as well. It doesn't take that long to learn and recognize the Kanji pieces.

The idea that Shogi and Chess would have different needs does seem rather inconsistent, as they are practically the same game. I have difficulty identifying any aspect of either game that would cause such a difference.

In theory they would be the same, but Shogi uses the drop rule, which by its nature necessitates a way to distinguish pieces that is not dependent on color.

Some years ago the 81doju server was also offering Chu Shogi, and I spent many hours there watching games, as my Chu Shogi program played there as a bot. One of the things I learned is that people tended to play at blitz speeds. Considering the much larger number of moves in a Chu Shogi game than in a Chess game, games of 30 min already require blitz speeds, and people don't want to play real-time games that last more than an hour.

Another thing I learned was that they tend to blunder because they overlooked an enemy piece that had ventured amongst their own pieces. (And then apologized "oh sorry, I had not seen that was your piece".) So it is not just a matter of the speed with which you identify the intruder; this directly correlates with the reliability with which you spot intruders. A deviating color sticks out, and attracts attention whenever it gets in your field of vision. Opposit orientation is only noticed when you are looking for it.

I did only play a handful of regular Shogi games in my entire life, but in one of those I beat a dan-rated player. Not because I outplayed him, but because he forfeited the game by 'nifu' (= dropping a Pawn in a file that already contained one). The other Pawn was hidden between my pieces. I don't think any player would make such an error if his pieces would have had a different color, but in the Japanese Shogi competition is appears to be a very common error. So much that it is considered part of the game there, and that the people running the 81dojo server did their utmost to prevent people from using their own client to connect, out of fear that such a client would provide 'computer help' by highlighting legal moves, and thus prevent forfeit by an illegal one. It seems the Japanese Shogi Association does not endorse Shogi servers where you cannot lose by an illegal move.

And what you are saying is basically that under easy conditions anything flies, and the negative impact of crummy equipment or other adverse conditions has less impact than in a stress test, where every advantage counts. Well, I cannot argue against that, but to judge the value of things you should test them under conditions that are sensitive to this value. That I can drive through Harlem at night in a tank, without experiencing any problem does not imply it is a safe neighborhood...

Anyway, nice chat. Good points from all parties. I appreciate.

In theory they would be the same, but Shogi uses the drop rule, which by its nature necessitates a way to distinguish pieces that is not dependent on color.

I did mention this as "an inconvenient necessity due to physical limitations in over-the-board play". And it is not entirely true; it is the combination of promotion and the drop rule that requires this. Otherwise you could have used different color on the backside. As this newly discovered 'South-African Chess' (Oatlali) does. (They replaced promotion by zone-dependent moving there.)

Personally I think it would have been better (in the sense that the player handling the pieces would play stronger) to flip pieces with differently colored sides on capture, and reorient them on promotion.

What is a better representation for playing a game is a verifyable fact, measurable by the average rating drop or raise people would get when having to play with one representation or another. E.g. representing all pieces by empty squares (aka blindfold chess) usually greatly degrades the rating of non-GM players, and can thus be considered an unsuitable representation for playing Chess.

Your example describes a handicap that would be universally detrimental for anyone playing against someone without this handicap. This points to some minimal requirements for piece design, namely that the pieces should be visible and clearly distinguishable. Both fully western sets and fully Japanese sets meet this requirement, and what makes the difference in how well someone can use one set or another will depend on more subjective, psychological differences, such as which set the person is more used to, or whether the pieces are written in a language one is able to read. Japanese people, and even Chinese and Korean people, will have an easier time with Kanji pieces than westerners will, because they can read the language the piece names are written in. Likewise, western players who are used to Staunton pieces will have an easier time recognizing pieces based on Staunton designs. To a lesser degree, western players may be less comfortable using wedge-shaped pieces than Japanese players would be. That is the preference you have indicated you have, though I have been playing with wedge-shaped pieces since I discovered Shogi in the late 90s and have no issue with them. So, my own preference is for pieces of a Staunton or my own Abstract design on wedge-shaped pieces.

And 'disagreeing with facts' is also known as 'being wrong'.

And what facts have you presented? You have not shown any data regarding which sets make it easier for people to play Shogi. The closest you have come to arguing for the superiority of western style Shogi pieces over Japanese style is this:

Playing Shogi across the board is an absolutely horrible experience. It is the main reason why the Dutch Chess Association has about 150,000 members, and the Dutch Shogi Association only 56...

While your data on membership counts may be correct, your explanation is not. This difference in membership counts is simply because the Netherlands is in Europe, where Chess is the dominant Chess variant, and not in Japan, where Shogi is. For similar reasons, you will find that Christian churches are more common in the Netherlands than Shinto shrines are.

For people that want to experience the difference between using pictogram and tile pieces, I set up a simple speed test cq dexterity game here.

So my response time for the pictograms is about half what it is for the Kanji tiles. But I'll point out that this is comparing two factors instead of just one. One is using either color or orientation, and the other is using Kanji or using pictures. Additionally, the Kanji pieces were at a smaller size, which made them harder to recognize, a more squarish shape, which made it harder to spot the orientation, and they included pieces not used in Shogi, which meant I could not always use my familiarity with the Kanji to recognize pieces.

The idea that Shogi and Chess would have different needs does seem rather inconsistent, as they are practically the same game. I have difficulty identifying any aspect of either game that would cause such a difference.

Two very big differences between Chess and Shogi account for this difference. In Shogi, pieces can switch sides, and promotion of each piece capable of it is to a fixed piece type. With wedge-shaped pieces that have different characters printed on each side, Shogi can easily be played with a single set of pieces. To play Shogi with western style pieces, you would need an extra set on hand, and each set would have to include separate pieces for the promoted and unpromoted versions of each piece. So, instead of requiring just 40 pieces, you would need up to 146 pieces if my calculations are correct.^* Playing Shogi with fully western style pieces has become feasible only with computer graphics, which do not require physical pieces to be kept on hand. While this may be the preferred way for some people to play online or against a computer program, people with experience playing it across the board will have had to get used to using wedge-shaped pieces.

But it is indeed good that people can choose their representation on a computer independently of that used by the opponent. If there is at least one suitable representation amongst the choices.

That's the important thing. Some will prefer the traditional Japanese style, some will prefer a fully western style, and some will prefer a hybrid.

* 40*2*2=160. 160 - 8 promoted Gold Generals - 4 promoted Kings - 2 normal Kings = 160 - 14 = 146.

Proof for who is right or wrong has to be collected. So far we have only seen that kanji tiles are no good for westerners. Would Japanese do better? I doubt it, but that is not proof. That the kanji are small is a feeble excuse: they must be smaller as a consequence of tiles being drawn around them. Which I claimed to be a bad idea, as it that unavoidably leaves less space. That there are now complaints about this smallnes sort of confirms that.

Upside-down pictograms would be better for westerners? Well, you can try that here. (Prepare for disaster!)

Funny to play with Bigorra. Yes, it takes much longer to spot an upside-down pictogram than to spot a wrong color. Neat. And the difference is much higher than with an upside-down kanji. Which would mean that it's easier to play with kanji than with oriented monochrome pictogram.

But I doubt of this conclusion. As a Westerner, I also need some time to decipher an upside-doown kanji whereas I need no time to decipher a pictogram (or almost).

It was not completely fair, as Bigorra used Archers, which look the same upside-down (or nearly so). I replaced them by Vaos now.

I guess the kanji were much easier, because at least the tiles all had the same shape, so you could search for the deviating shape you expected. The pictograms are all totally different (which is of course good for recognizing type rather than side, but we don't test that here). I therefore suspect that the kanji, which are also all different, are not much help even to those that can read them, and that they would mainly look at the tile shape.

Distinguishing the upside-down pieces is a little easier than being completely blindfolded. But only a little...

pictograms 18 1.831 kanji tiles 17 17.078

It feels like it's easier to spot orientation when all the pieces have the wedge shape than if they're all differently shaped pictures. Couldn't a similar point be made regarding 3D piece descriptions (which jocly uses) as opposed to the 2D alternatives?

The Jocly 3D pieces are kanji tiles, and for the Shogi variants are little more than a flat diagram viewed in perspective. So I guess what holds for the 2D representations pretty much holds for 3D as well: for recognizing the side a piece is on you go mostly by the tile orientation, but you will only notice that when you consciously focus your attention to it.

Even when the "spot the intruder" game had not been explained, a person that is shown a case that uses colors would almost instantly and spontaneously remark upon a wrong piece being amongst the army.