Talk:Shogi strategy

Anaguma diagram?

I don't think the diagram for Anaguma castle is standard. That's certainly a structure you want to take when your opponent's rook is bearing down on your castle, but the most basic form of Anaguma places the bottom gold on 3i, not 4h. (That way, 2h silver puts string to one of the golds for extra protection in the endgame.)

Considering this is an article for people who are just starting shogi or who do not play shogi, I feel as though the standard variations of castles should be displayed. Hirohiigo (talk) 03:32, 23 September 2009 (UTC)

Optional step #6

"(Optional) Move the fourth file pawn one square forward. This makes it harder for the most direct threat-mate on the castle to be made." Is that fourth from the left (Western) or fourth from the right (Japanese)? OneWeirdDude (talk) 17:34, 11 June 2013 (UTC)

OneWeirdDude, in this case, it's moving the pawn on 4g to 4f. This allows the the gold to move from 5h to 4g to move from a regular Mino castle to a High Mino (高美濃, taka minō) which is better against attacks from the front and also defends the 3f square which is one of the weak points of the Mino. The diagram is a little confusing because the rook on 6h is where it goes in the 4th file rook (四間飛車, shikenbisha) opening. That cannot be the fourth file for this particular diagram though because that pawn is actually further away from the king and has nothing to do at all with the strengthening of the castle. Letters and numbers should be added to the castle diagrams to clearly identify the rows and columns so that there's no chance for such confusion. -Marchjuly (talk) 15:45, 9 March 2014 (UTC)

Mino castle: step 5

The first part of the Step 5 Move the edge pawn one square forward is almost always considered to be a good move. The same cannot be said about the second part two is even better. In fact, moving that lance pawn one additional square forward often is quite bad because it actually serves to speed up your opponent's attack. It also creates a weakness on the 1-f square which can now be used for a piece drop to either give check (e.g, a knight drop) or threaten the 2-g and 1-g squares (e.g., a gold or silver drop). Saying two is better is too strong of a statement and should be removed. It may sometimes be good, but at the same time it may sometimes be bad. This is why you don't see such a move played too often in professional/high level games unless it leads to a clear advantage. - Marchjuly (talk) 15:22, 9 March 2014 (UTC)

Mino castle: step 1

Moving the rook to the left side of the board is the first step, but there is no reason it absolutely has to be moved to the 6th file. 4th File Rook (四間飛車, shikenbisha) is just one type of ranging rook opening. There's also Central Rook (中飛車, nakabisha), 3rd File Rook (三間飛車, sangenbisha) and Opposing Rook (向い飛車, mukaibisha). All of these are played at all levels of shogi (including pro), so whether one type is preferred to another is completely subjective and depends on individual playing style more than anything else. - Marchjuly (talk) 23:40, 21 March 2014 (UTC)

Mino castle: step 6

This makes it harder for the most direct threat-mate on the castle to be made. is another fairly subjective statement. In its basic form the Mino castle is quite strong against attacks from the opposing side. It is, however, not very strong against attacks from the front. Moving the 4th file pawn one square forward allows the castle to develop into a High Mino (高美濃, takaminō) and then possibly even further into a Silver Crown (Mino) (銀冠, ginkanmuri). Developing the castle this way strengthens it against attacks from the front, increases the scope of the defending pieces by allowing them to exert some control over the center, and makes it easier for the Mino side to shift from defense to attack if a chance presents itself. In the basic Mino, the 3f square is very weak, so moving into a High Mino helps defend this square against not only mating attacks, but attacks on the head of the knight after it has been moved to 3g. Moving this pawn is also necessary in the Central Rook opening because the rook is on the middle (5th) file so there is no place for the gold on 6h to go other than 4g. - Marchjuly (talk) 00:05, 22 March 2014 (UTC)

Mino castle: general note

I think whoever edited this last got rook and bishop mixed up--unless they're playing a version in which you begin with the rook on the right. I learned with rook on the left and bishop on the right, but I know it varies. Perhaps add a word or two of clarification. — Preceding unsigned comment added by 74.90.244.255 (talk) 21:09, 13 September 2014 (UTC)

Actually 74.90.244.255, the starting position for the rook is always the square 2h, or 28, for sente (i.e., black) and always the square 8b, or 82, for gote (i.e., white) and both of these are on the "right" side of a standard shogi board from the viewpoint of the player whose piece the rook is. That is why the expression "move to the left" is used. Perhaps when you say "left", you are looking across the board and referring to your opponent's rook starting position from your perspective. That indeed would be "left", but such usage would not be a appropriate here. - Marchjuly (talk) 23:36, 13 September 2014 (UTC)

Castle diagrams

Just a formating suggestion, but it might be better use Template:Shogi diagram for the examples of the various castles. The template can be tweaked to display the ranks and files in different ways or to only show the relevant part of the board. The template could be used as follows for the Gold Yagura.

 {{shogi diagram img||金矢倉 Gold Yagura<br /><small>後手 White</small>
 |    |    |    |    |    |    |    |    |
 |    |    |    |    |    |    |    |    |
 |    |    |    |    |    |    |    |    |
 |    |    |    |    |    |    |    |    |
 |    |    |    |    |    |    |    |    |
 |    |    | ps | ps | ps |    |    |    |
 | ps | ps | ss | gs |    |    |    |    |
 |    | ks | gs |    |    |    |    |    |
 | ls | ns |    |    |    |    |    |    |
 |先手 Black}}
 

which would look like this

{{shogi diagram img||金矢倉 Gold Yagura <br /><small>後手 White</small>
|    |    |    |    |    |    |    |    |
|    |    |    |    |    |    |    |    |
|    |    |    |    |    |    |    |    |
|    |    |    |    |    |    |    |    |
|    |    |    |    |    |    |    |    |
|    |    | ps | ps | ps |    |    |    |
| ps | ps | ss | gs |    |    |    |    |
|    | ks | gs |    |    |    |    |    |
| ls | ns |    |    |    |    |    |    |
| 先手 Black}}

I also think it might be best to use a simple column markup to add the diagrams. This would make it easier to keep them in or at least close to the relevant section were they are being discussed. For example, the Yagura castles could be added like this

 {{Col-begin|width=50%}}
 {{Col-2}}
 {| class="wikitable"
 |+ 金矢倉 Gold Yagura
 ! ９	!! ８ 	!! ７ 	!! ６ 	!! ５ 	!! ４ 	!! ３ 	!! ２ 	!! １ 	!! 
 |-
 |   	||  	||  	||  	||  	||  	||  	||  	||  	<th>一 a</th>
 |-
 |   	||   	||   	||   	||   	||   	||   	||   	||   	<th>二 b</th>
 |-
 |   	||   	||   	||   	||   	||   	||   	||   	||   	<th>三 c</th>
 |-
 |   	||   	||   	||   	||   	||   	||   	||   	||   	<th>四 d</th>
 |-
 |   	||   	||   	||   	||   	||   	||   	||   	||   	<th>五 e</th>
 |-
 |   	||   	|| 歩	|| 歩	|| 歩	||   	||   	||   	||   	<th>六 f</th>
 |-
 | 歩	|| 歩	|| 銀	|| 金	||   	||   	||   	||   	||   	<th>七 g</th>
 |-
 |   	|| 玉	|| 金	||  	||   	||   	||   	||   	||   	<th>八 h</th>
 |-
 | 香	|| 桂	||  	||   	||   	||   	||   	||   	||   	<th>九 i</th>
 |}
 {{Col-2}}
 {| class="wikitable"
 |+ カニ Crab
 ! ９	!! ８ 	!! ７ 	!! ６ 	!! ５ 	!! ４ 	!! ３ 	!! ２ 	!! １ 	!! 
 |-
 |   	||  	||  	||  	||  	||  	||  	||  	||  	<th>一 a</th>
 |-
 |   	||   	||   	||   	||   	||   	||   	||   	||   	<th>二 b</th>
 |-
 |   	||   	||   	||   	||   	||   	||   	||   	||   	<th>三 c</th>
 |-
 |   	||   	||   	||   	||   	||   	||   	||   	||   	<th>四 d</th>
 |-
 |   	||   	||   	||   	||   	||   	||   	||   	||   	<th>五 e</th>
 |-
 |   	||   	|| 歩	||  	|| 歩	||   	||   	||   	||   	<th>六 f</th>
 |-
 | 歩	|| 歩	||  	|| 歩	||   	||   	||   	||   	||   	<th>七 g</th>
 |-
 |   	|| 角	|| 金	|| 銀	|| 金	||   	||   	||   	||   	<th>八 h</th>
 |-
 | 香	|| 桂	||  	|| 玉	||   	||   	||   	||   	||   	<th>九 i</th>
 |}
 {{Col-end}}

which would look like the following in the article

金矢倉 Gold Yagura

９	８	７	６	５	４	３	２	１
									一 a
									二 b
									三 c
									四 d
									五 e
		歩	歩	歩					六 f
歩	歩	銀	金						七 g
	玉	金							八 h
香	桂								九 i

カニ Crab

９	８	７	６	５	４	３	２	１
									一 a
									二 b
									三 c
									四 d
									五 e
		歩		歩					六 f
歩	歩		歩						七 g
	角	金	銀	金					八 h
香	桂		玉						九 i

The widths of the columns can be easily adjusted depending upon the their total number. -- Marchjuly (talk) 23:33, 28 April 2016 (UTC)

Makes sense. My CSS & wikisyntax-fu has atrophied. Wasn't sure how order should be imposed on the table positioning. Generally, I think that floats flow better. But, that's only when there's enough text surrounding them unlike this article where it's probably going to be the case that there's plenty of castles without text explaining them leading to tables doing bumper cars with each other.

Although I like text in tables more than images due to their crisper edges and WYSISWYG-ness, the template is easier for turning characters upside down. And, it's cross-wiki and thus a generalized wikipedia solution. – ishwar  (speak) 00:06, 29 April 2016 (UTC)

Here's a question: is there a way to highlight a cell via different text or color or something?— Preceding unsigned comment added by Ish ishwar (talk • contribs) 09:23, 29 April 2016 (UTC+9)

@Ish ishwar: Sorry for not responding sooner. There is a way to highlight a specific piece, but I think you've figured that out. There may be a way to highlight a specific square, but that probably would have to be added to the template's primary mark up. I also see you figured out how to further incoporate the diagrams into the text using "divisions". Article looks much better than before, but I think some footnotes at key places might be a good idea. I know that the new content you added to be true, but the article is there for anyone, including those who know nothing about shogi, to read/edit so some of might be tagged with {{citation needed}} or worst case removed per WP:OR. There are tons of Japanese books which could be cited for any one castle or opening, but Hoskings' The Art of Shogi pages 43-45 could be used for "ranging" rook vs. "static rook" and pages 46-55 could be used for the various castles. Also, I think "bōgin" and "jōseki" are technically the preferred way to romanize those words per MOS:JAPAN#Romanization, but the terms are often written without the long-vowel in the English books/websites so I think "bogin" and "joseki" would be fine. -- Marchjuly (talk) 20:35, 29 April 2016 (UTC)

It's fine. I've taken longer than a year to respond to folks. Plus, the article hasn't changed much in a year anyway. No hurry. Yeah, I snooped around the code stuff as I sometimes do. I don't have that book, so maybe you can document the undocumented castles if they're in there. I just thought it was funny that article namedropped a bunch of castles without any picture or description about what these castles are. (Similarly, the 'opening' section doesn't really discuss a single opening besides the namedrop, which is disappointing.) English doesn't have long vowels or macrons, so using just o is sensible. However, they aren't really English words, so I didn't see there being any standard from usage. But, I'll believe you. O is fine.

Here's another question. I see the term Side Pawn Picker used all over the internet. Some of this is obviously from that Youtube dude's translation. But, how is 取り as in 横歩取り usually translated into English in shogi books? "Pick" seems to be a weird translation in the context of a chess game. I used "capture" especially since "capture" is the translation in other shogi (and, of course, chess) contexts. (Does one ever "pick" pieces in western chess?) – ishwar  (speak) 06:02, 30 April 2016 (UTC)

I've seen "side pawn capture" before in Habu's Words and Hosking refers to it as simply "side pawn" () in the Art of Shogi, but I've actually never seen "side pawn picker" until now. It sounds like an original translation to me. I think lots of the videos, etc. that are posted online in English are helpful (especially for beginners), but I do think some of the translations are a bit too "original". Personally, I think it's much better to use the Japanese names for the openings/castles/pieces, etc. as much as possible although I do understand the desire to try and make things more accessible to non-Japanese players. I'm probably a bit out of touch with the way they refer to things these days since I usually stick to Japanese materials when it comes to shogi, but I do agree though that "capture" sounds more natural. -- Marchjuly (talk) 07:53, 30 April 2016 (UTC)

Ok, good. I'm glad that's not a standard translation. (Although if his videos are influential enough over time, it might become standardized...) I love his videos though. On the names, I think translations are actually best because while Japanese words have meaning for Japanese speakers they don't have any meaning in English. So, using only Japanese transliterations divorces any mnemonic value that the Japanese word have. However, some translations are poor and the mnemonicity is lost. (For example, common climbing silver is too vague in that it only indicates upward movement whereas the Japanese term indicates a straight upward trajectory, a better visual metaphor.) But, I do guess the reasoning behind your preference is the current poor state of shogi literature in English. One is really forced to read Japanese for advanced study as nothing is available in other languages. So, it's important to have a reference for the Japanese terminology. I suggest that wikiland use both as much as possible. – ishwar  (speak) 01:13, 1 May 2016 (UTC)

@Ish ishwar: My personal preference for naming is just that. If the consensus is to use both English and Japanese, then that is fine with me. Anyway, there is another way to link to Japanese Wikipedia articles using Template:Link-interwiki. It allows a red-link to be created in a way which does not show the target link for the Japanese Wikipedia article. Thhis be used in combination with the "nihongo" or "eigo" templates. For example, {{nihongo4|{{link-interwiki|vertical-align=sup|en=Double Wing Attack|lang=ja|lang_title=相掛かり}}|相掛かり|aigakari}} will look like this Double Wing Attack (相掛かり, aigakari) to the reader. If someday someonce creates an article for "Double Wing Attack", the red link and the link to Japanese Wikipedia will automatically be removed. Finally, one last suggestion. It might help to add Template:In use or Template:Under construction to the article. These just sometimes help to prevents edit conficts/unnecessary reverts since it lets others know that there is an editor working on expanding the article or currently working on the article. I plan on using "In use" when I add citations since I will probably add them in bunches to save a bit of time. -- Marchjuly (talk) 01:25, 2 May 2016 (UTC)

That's clever. (Wonder when that was invented. Somewhat more onerous for the editor though.) It makes more sense. Going to Japanese wiki is probably not the reader's expectation. Is nihongo4 still necessary? Web browsers/operating systems still aren't multilingually up-to-date in 2016?

If you want to use 'in use', that's fine. Wikipedia now seems to warn you if this happening though. (Guess it's not perfect? Which is why you suggest this?) – ishwar  (speak) 16:33, 2 May 2016 (UTC)

I'm not sure when they came up with the "link-interwiki" template, but it's been around for at least as long as me. I also agree that it is a little hard to use; I keep forgeting how to mark it up so it always takes a few tries to get it right. It does create lots of redlinks which seem to annoy some editors, especially when there's zero or very little prospects of them ever becoming stand alone articles. Anyway, I only suggested it as one more option to consider.

Also, not sure about the "nihongo4" template. Maybe as you say it's really no longer needed and the romaji kanji can simply be inputted. For reference, there are three other "nihongo" templates as well as an "eigo" template and each does something slightly different. "Nihongo4" doesn't add a link explaining Japanese character sets.

Finally, "In use" was just a suggestion, mainly because yesterday or the day before we were both editing at the same time which created an edit conflict. It's probably only really needed when you're planning on editing the entire article (i.e., opening the edit window for the entire article) and not just a specific section since anyone else trying to edit at the same time is going to get hit with an edit conflict. -- Marchjuly (talk) 22:50, 2 May 2016 (UTC)

Change in citation style

I was bold and changed the citation style a bit with this edit. My main reason for doing so is that the article has been expanded quite a lot recently (in a good way in my opinion), but much of the new content is not supported by inline citations. Much of what has been added can be found in books by Hoskings and Fairbairn and I will gradually be trying to add supporting sources as I can. I switched to short footnotes simply because it makes it a little easier to format multiple citations of different pages from the same source, than by using the simple "ref name" method. It also allows the two books being primarily cited in the article to be differentiated from the more general references listed in the "Bibliography" section, but not directly cited. Fairbairn has been cited 10 times, but no specific page numbers are given, so I am going to try to track them down which may make the reasons why short footnotes are a bit better a little more clearer. -- Marchjuly (talk) 00:56, 2 May 2016 (UTC)

Rook position

@Ish ishwar: As you probably well know, the way files are referred to when discussing the position of the rook in various openings can be somewhat confusing even in Japanese. For example, "shikenbisha" means "4th file rook", but from blacks perspective the rook is actually on the 6th file (6h). Same goes for "sangenbisha" which means "3rd file rook", but rook is actually on the 7th file (7h). Occasionally this "unusualness" is briefly discussed by pros during their game commentary or in books, but most of the time (at least In Japan) it's just assumed the reader/audience understands these particular usages. I've heard a few pros say "rokken bisha" ("6th file rook"), but pretty much only when they are in uber explanatory mode and almost always just to clarify things for beginners. If the rook is simply moved to the 4th file (4 suji) then that is expressed without any problem; it only seems confusing when it comes to the names of openings. The only one which seems problem free is "nakabisha" which means "central rook" and is the same no matter whether you count the files from left to right or vice versa. The only reason I'm bringing this up is that the names you've given for the diagrams for "Static rook" and "Rangin rook" might be a little confusing to someone not familiar with this particular naming quirk. When "4th file rook opening" is used in English books or on English websites, it almost always refers to "Shikenbisha" with the rook moved to 6h (not 4h); the same goes for "third file rook" with regards 7h (not 3h). For "mukai bisha" "opposing rook" is almost always used and not "8th file rook" even though the rook is actually on 8h from black's perspective. It might be helpful to add the rook for white to the diagrams and a bit of explanation about the naming. -- Marchjuly (talk) 23:12, 2 May 2016 (UTC)

Yeah, I've thought about that. My nebulous masterplan thinking was something like: (1) distinguish between coordinates vs nomeclature, (2) coordinates used in general description and nomenclature with explanation of nomenclature used in individual joseki description. (Because from black's perspective, the 7th file is the 7th file regardless of what it's called in a joseki.) However, I don't know if that's best. It definitely is the case that the terminology is confusing. (I know I found it confusing when I first was trying to figure this stuff out.) Part of this confusion is that these 4th/3rd/opposing file terms are stemming the fact that they are often (or originated?) played by white (even though sometimes they are played by black too!) and the convention is to always display diagram coordinates with black on the bottom/white on the top. One solution is yours: display these rook positions from the perspective of white. The demerit of this is that even if you are white, you don't play upside down so it looks weird to see it that way. Another solution is put white on the bottom and black on the top so that one can see it as one would play it. The demerit of that is that the diagram convention is now flouted perhaps confusingly.

Of course, one could do both. The merit of doing both is that a player needs to be able to recognize rightsideup and upside positions anyway. The demerit is that it kind of creates diagram soup. I don't have problems with soups (perhaps this is becoming apparent to you in my editing) but some might. And, I'd also point out to potential objectors that wikipedia don't have the same economic space constraints that the print medium has (although this doesn't mean readability should ever be sacrificed for the sake of diagram explosion).

As it stands presently, the diagrams and classification and the opening names clash. So, there needs to be some description somewhere at some point.

(You know, you can always add your points to the article. You probably know shogi better than me anyway.) – ishwar  (speak) 23:52, 2 May 2016 (UTC)

No worries. Just for reference, I often do shoot first and talk later when I edit. However, it seems that you have a certain plan in mind as part of this expansion and so far it's looking good. So, I do not want to change too many things until it all seems in place. My last post also wasn't clear about the "white rook" part. I'm not suggesting that the diagrams be flipped so that white is on the bottom; I just think it might be a little more easier to understand to add the white rook on its original square to the existing diagrams when discussing openings, especially for the "opposing rook" diagram. Instead of two diagrams, it may be possible to tweak the template so that the upper half is for white and the lower half is for black. For example,

{{shogi diagram img||'''Ranging Rook'''<br />4th File rook for White (Rook moved from 8b to 4b)<br /> <small>△</small>
 | lg | ng | sg | gg | kg | gg | sg | ng | lg   
 |    |    |    |    |    | rgl|    | bg |    
 | pg | pg | pg | pg | pg |    |    | pg | pg  
 |    |    |    |    |    | pg | pg |    |    
 | yy | yy | yy | yy | yy | yy | yy | yy | yy  
 |    |    | ps | ps |    |    |    |    |    
 | ps | ps |    |    | ps | ps | ps | ps | ps 
 |    | bs |    | rsl|    |    |    |    |    
 | ls | ns | ss | gs | ks | gs | ss | ns | ls 
 |<small>▲</small><br /> 4th File Rook for Black (Rook moved from 2h to 6h)}}

I've used the yellow just for this example, but maybe there's a way to null out the 5th rank from the diagram so that the break between top and bottom is clearer. It would be an interesting expansion of the template and is a trick often used in shogi books to save space. -- Marchjuly (talk) 00:53, 3 May 2016 (UTC)

That's a good idea. But, I'll think you'll need to have a new template (or add in some conditional formatting with a formatting toggle to do that. The current template expects that table structure with exactly those 84 arguments.

Yeah, maybe it was me that wasnt clear. I got what you wrote. The reversal of black = bottom & white = top was my idea. I meant that if you show:

{{shogi diagram img||

| lg | ng | sg | gg | kg | gg | sg | ng | lg   
|    |    |    |    |    | rgl|    | bg |    
| pg | pg | pg | pg | pg |    |    | pg | pg  
|    |    |    |    |    | pg | pg |    |    
|    |    |    |    |    |    |    |    |    
|    |    |    |    |    |    |    |    |    
|    |    |    |    |    |    |    |    |    
|    |    |    |    |    |    |    |    |    
|    |    |    |    |    |    |    |    |    

|}}

then, it's a little bit jarring to view it upside down if one is trying to know what it looks like when one plays it. I mean that diagram's flawlessly informative of course: the information is all there. But, when trying to recreate it for a player's perspective, one has to either flip it in one's head or flip it on a board/app. You can argue that this is more of a pedagogical point. And, someone can drop a line that wikipedia isn't a textbook. I don't necessarily agree with that. I hope I'm being clearer now. (Not to say i'm right though!) – ishwar  (speak) 03:04, 3 May 2016 (UTC)

Relatedly, I thought about this for castles, too. It's nice to see the castle upside -- the way it looks if black is looking at white's castle. So, that's helpful for the reader. The trade-off is that the diagram then has more information. And, it's always easier to understand things when one strips away extraneous information. – ishwar  (speak) 03:37, 3 May 2016 (UTC)

Yagura castle diagrams

The "right-left" layout for the Yagura castles does look similar to what is often done in shogi books, but it might be a little too ambitious for Wikipedia. The left-floating diagrams of the "Silver Yagura" and "Half Yagura" were causing the "Yagura" and "Mino" sections to overlap a bit as seen here in this old version. I tried to fix this by manually inserting line breaks between the sections, but this is probably not a good way to solve this problem. Another problem is that there is really no margin between the two left diagrams and the text discussing them. text and diagrams seem to meld together which could make the text difficult to read on some monitors. One possible way to fix both problems at once would simply to be to move all 4 Yagura diagrams to the end of the section much like a gallery of images. This would also have the benefit of making it possible to add more diagrams or text without one affecting the other. I'm not technically well-versed in wiki/html markup, so if there's a better alternative, then please discuss. -- Marchjuly (talk) 05:02, 9 May 2016 (UTC)

I wasn't seeing that. But, I do see that when I shrink my font down 75%. I think there's some solution to that. – ishwar  (speak) 06:14, 9 May 2016 (UTC)

Thanks for taking another crack at it. Whatever you did seems to have solved both problems and I am no longer seeing any overlap. -- 06:27, 9 May 2016 (UTC)

: Ok. Solution is to add <br clear="all"> after the text right before the start of the next section. (This is the old way. Somebody would need to look up the modern CSS way to it. But, regardless, the old way works -- on my browsers anyway...) – ishwar  (speak) 06:28, 9 May 2016 (UTC)

Kyokumenpedia as a reliable source

I think Kyokumenpedia is probably acceptable as an external link per WP:ELYES, but I'm not sure if we can use it as a reliable source per WP:UGC. It looks to me like it is an open wiki from its terms of use page that anyone can register for and edit, and also that it does not have any form of editorial oversight other than what you'd find on a typical open wiki. I think there are better sources for showing the frequency of opening use in various published books, etc. For example, I think such information can be found in the various yearly almanacs () like these published by the JSA. The JSA also lists the results of all official games between professionals or involving professionals on it's official page by month all the way back to January 2011 and older results are archived here. These listing don't necessarily mention specific opening, but some of them include links to other pages where the full game score might be able to be seen. Personally, I think the info in Kyokumenpedia is probably pretty accurate, since I don't envision much shogi related vandalism occurring on a site like that, but once again I'm just not sure we can use it as a reliable source even if it's explicitly attributed by name in text. -- Marchjuly (talk) 00:53, 10 May 2016 (UTC)

One can edit wiki-related text, but I don't think the game records are actually editable by the public. In fact, one can't even see the which professionals played which game, only the "superuser" has these viewing rights. So, an objection based on the wiki aspect doesn't seem relevant.

A more cogent objection would be that one mistrusts the veracity or accuracy of the site administrator(s) or the game records themselves. For example, I don't see what the game record sources are. I don't see what the time period of game recording sampling is (presumably it doesn't back hundreds of years), and so on. Tomohide Kawasaki is one administrator. He's the Hidetchi youtube shogi promoter, 81Dojo director, and translator/coauthor of some Madoka Kitao books. I don't know if he's the only one. Although Kawasaki is connected to the JSA via 81Dojo, this site doesn't seem to be related to the JSA. – ishwar  (speak) 01:39, 10 May 2016 (UTC)

The question then might whether it is appropriate to consider Kawasaki and "expert on shogi" per WP:UGC for Wikipedia's purposes. As I said, I do not feel there is any possibility of intentional misrepresentation going on, but my opinion is just my opinion. I am not sure how others would feel if this was brought up for discussion at WP:RSN by someone else (perhaps someone with less familiarity with shogi). Perhaps I am trying too hard to see smoke where there is currently no fire and being more cautious than needed. -- Marchjuly (talk) 02:37, 10 May 2016 (UTC)

I don't know if he's an 'expert'. Is Fairbairn an expert? Or Trevor Leggett? They have written/translated books. Also, how does the 'expert' label apply here? Isn't the more applicable thing here whether Kawasaki is an expert in database management, coding abilities, and game record procurement since we're not relying on his shogi knowledge here? We're just counting games that meet some formal criteria for the first two moves, which a quantitatively minded person can do without any knowledge of shogi at all. So, again, I'd say the more relevant concern is whether the database itself is done correctly and also how representative the database is of the average professional shogi game. – ishwar  (speak) 17:35, 10 May 2016 (UTC)

Yes, you make good points. "Expert" was just used to encompass all of the things you mentioned as well. As long as it can be shown that the game scores are in fact "the game scores" and that they come from professional games, then using it in such a way should be OK much in the same way a database of academic articles is considered OK. Kawasaki would really only need to be "a shogi expert" if we wished to cite his analysis or opinions on openings or other shogi-related matters. -- Marchjuly (talk) 22:13, 10 May 2016 (UTC)

It's easy to confirm individual games. For example, http://kyokumen.jp/positions/4037306 is the 42nd move of http://live.shogi.or.jp/asahi/kifu/6/asahi201301170501.html. Manually inspecting thousands of games is onerous though. – ishwar  (speak) 00:06, 11 May 2016 (UTC)

Understand completely. For what it's worth, I'm fine using it as a RS as long as it's restricted to statistical content and game score accuracy is not really an issue. If that's the consensus, then I have no problem with the unhidding of that particular bit of text. A link to this discussion can be added as part of the edit sum "See Talk:Shogi strategy and tactics#Kyokumenpedia as a reliable source" for reference. If anyone else has concerns about the source, then at least they'll be aware of this discussion and then can comment if they like. Does that sound workable to you? -- Marchjuly (talk) 01:00, 11 May 2016 (UTC)

yep – ishwar  (speak) 02:16, 11 May 2016 (UTC)

tables

One does not use HTML for tables. Use wikicode. It was just added by you, so do it correctly. You've been around here long enough to know the right way. You should also know to do it right the first time. Bgwhite (talk) 04:26, 13 May 2016 (UTC)

Hi Bgwhite. I saw the edit you made and I realize your comment is probably not directed to me, but I did create Template:Shogi diagram so if there's a problem with it then I'd like to know. It was the first and so far only template I've created (I got the basic syntax from Japanese Wikipedia) and templates in general are not something I know a lot about. It's been a bit of a "figure things out on the fly" for me so any, suggestions you may have on ways to further improve it would be most appreciated. Thanks in advance. -- Marchjuly (talk) 05:00, 13 May 2016 (UTC)

Currently, the wikicode for tables doesn't work within the template. Bgwhite, you're welcome to fix it, of course, especially if you are knowledgeable about html & wikicode & tables (as you might be since you initiated this deletion). I'm not sure what the problem is yet – right the first time. But, your current edit is just removing information, which seems contrary to purpose of working on an encyclopedia? – ishwar  (speak) 05:03, 13 May 2016 (UTC)

yagura/fortress?

so, older translations (fairbairn, hosking) use the translation fortress while wikipedia and other newer things use the untranslated japanese word yagura. Which to use? – ishwar  (speak) 21:30, 18 May 2016 (UTC)

I personally prefer yagura, but I think back in the day (pre-Internet or early-Internet) everything tended to be translated to make it easier to understand and more "western" sounding. However, I think these days that might no longer be such an issue, especially since there are quite a number of Japanese words (karaoke, pokemon, sushi, manga, anime, etc.) which are pretty much mainstream. Maybe all that is needed is a single explanation that yagura means "fortress" when the term is first introduced and then use the Japanese term thereafter. For example, in the non-Japanese shogi world (including on Wikipedia), dan and kyu are used to refer to player ability. Also, major titles and title holders are referred to using their Japanese names. Yoshiharu Habu is referred to as Habu Meijin, not really as Habu Grand Champion. Akira Watanabe is referred to as Watanabe Ryu-oh, not Watanabe Dragon King, unless someone is really trying to be literal for a reason. Even shogi is referred to as "shogi" and not as "Japanese chess". At some point, the term, if used enough, will stop being a Japanese word and becomes a topic-specific word. Still, if the consensus is to use "fortress" for the castle and "yagura" for the opening, then that is fine with me. We could, however, also write "yagura castle" and "yagura opening" when it is necessary to make such a distinction. -- Marchjuly (talk) 01:25, 19 May 2016 (UTC)

I don't think 'western sounding' is a consideration. It's just whether a term is established in the lexicon (vocab) of an average English speaker. The English language (or their speakers) don't have any problem borrowing words from other languages (in contrast to a language like Icelandic, for example).

The question is what to do when dealing with terms in the source language that aren't established in English. This is always the case.

The terms you mention like dan, Meijin, etc., I'm not sure if they are usually translated instead of borrowed. So, those are non-issues, i think.

The fact that some folks online are not using a translation for yagura makes me think that following them is the way to go. But, I'm not really sure what state of the English speaking world is. (I also note some Spanish websites – and maybe Portuguese if i remember right – also introduce yagura as well as anaguma. This might be due to influence from the English language websites.) Maybe this + you = consensus? (And, the old books being outvoted by recency.)

I don't like using fortress as castle and yagura as opening. But, my objection is based on just the fact that i know the word yagura is used for both in Japanese. (Which seems like an invalid objection, really.) Another possible objection is that these castles and openings (although not obligatorily connected) go together like red beans & rice (or liver & onions or corn beef & cabbage or ...). So, it's useful to connect them via terminology? I don't know.

At any rate, i'm leaving it with yagura as the wiki name with a single mention of fortress as an alternate equivalent. – ishwar  (speak) 02:24, 19 May 2016 (UTC)

Buttocks-Gold Checkmate

@Ish ishwar: You've been doing a really nice job on not only this article, but the other articles you've spun off from this. I've been trying to tweak little things here and there, but to be honest, there's not a lot that needs fixing (at least shogi wise). I did, however, notice that this is not mate because the silver on 8b is not protected. Any piece that defends the silver on 8b can be added to the diagram to make it a mate, so I'll leave it up to you to tweak it as you see fit. -- Marchjuly (talk) 07:37, 27 May 2016 (UTC)

Ah, good catch. That should've been a bishop, I guess. For this tactical stuff, rather than copying examples directly from books, I've been tweaking the examples to be somewhat different but in the same spirit. (My thought is to detract from book authors as little as possible so one can still find value in their examples.) That may result in errors like this. So, proofreading needed. For the josekis, they're just josekis, so no room for originality. Of course, there will probably be some typos and mistranspositions there, too. – ishwar  (speak) 07:56, 27 May 2016 (UTC)

Adding the bishop like you did makes it mate, but it also means there can be a mate on 8b with the gold as well. What about adding rook for gote on the b-rank? This would mean that dropping the gold on 9a would be the only way to mate. Placing the rook anywhere other than 8b on the b-rank should do the trick. You can also you another gote piece on 8a to cover 8b if you like, but it seems that the point of the example is to show that gote's king cannot go to 8a which would be impossible if say a lance or silver was on that square. I guess you could try a bishop or silver on 7a if you want. Also, neither sente nor gote having pieces in hand does not affect the solution, but often in tsumi problems, which this is sort of like, gote will be given 残り or all the remaining pieces in hand even though that cannot use any of them. Do you think that is worth tweaking? -- Marchjuly (talk) 08:54, 27 May 2016 (UTC)

Good point. I think the example should ideally have the gold attacking all the possible escape routes of the king as well as attacking the king itself. And, the example wasn't that due to the sente bishop. So, I used a bishop on the a rank as you suggested and put a nondefending knight on the side of the king blocking the king's side movement, then the (defended) gold and only the gold delivers the mate. Good. – ishwar  (speak) 20:12, 27 May 2016 (UTC)

Dancing Pawns

The way the diagrams are used to show this is nice. However, I think three pawns in hand at the beginning are needed to make it work right. Also, in the final position the gote gold can still escape to 3d, so it's not a lost piece. For sure, this still allows sente to create a tokin on 2c which is typically a big advantage. I'm not sure whether more diagrams would be overkill for that section. Any opinions either way? -- Marchjuly (talk) 07:51, 27 May 2016 (UTC)

You're right. I lost track of the count.

That's true about the gold escape. I was trying to simplify a typical example that used a gold-silver-gold sequence by using a bishop-gold sequence. I think the solution is add a pawn on the side of sente's silver. Then, the gold will captured and there's still the undesirable gold-pawn exchange.

What do you mean about more diagrams? (And, to be honest, I've wondered whether there's already too many diagrams here...) – ishwar  (speak) 08:13, 27 May 2016 (UTC)

I wasn't sure if the point was to show how the gold could be "trapped" or how a pawn promotion could be forced. If the former, then adding the pawn as you suggested is all that's needed. If the latter, then maybe two more diagrams would be needed: one for G-3d and one for P-2c+. I like the diagrams because I like shogi books, but I am a little bit worried that other editors (who might not be that into shogi) might start this as being a bit of overkill. I like photos too, but lots of editors feel a little goes a long way when it comes to images and that Wikipedia articles should be mainly text and not image galleries. I'm not criticizing, just trying to anticipate how others might see it. There is also page loading time to consider. For awhile there, I did notice that the page was taking a bit longer than usual to load. Maybe that's really not an issue now that the other articles have been created. -- Marchjuly (talk) 08:42, 27 May 2016 (UTC)

I thought the point was that even if a gold can dance it can't dance its way out of this situation. That seems to be the focus in the examples. Adding the pawn like i did makes it more similar formally to the 金銀金-type example (e.g. [1]). In those examples, if the gold tries to escape, then it will also get nabbed by sente pawn. However, not all blog stuff actually has that feature. For instance, they include a king along with the gold. Then, the reasoning seems to be that the gold shouldn't try to escape and abandon the king.

So, maybe the pawn promotion is a separate point?

As for the abundance of pics, it's only a problem if there's a valid objection. A preference of text over a visualization applied without thought doesn't make any sense to me. One tradeoff is between showing moves to the reader visually vs asking the reader to visualize the moves in their head and saving space. Since I'm viewing wikipedia space as 'costless', an objection along these lines is unconvincing. Too many pics may an make article hard to read, so an objection in these terms may be a good one. I don't think we're at that point yet, but it's perhaps subjective. I can't characterize the images here as a gallery since we are not displaying art or showing stuff to see just to be seeing it. (At least, i don't think we are?)

It's always possible to reduce the diagrams by cutting out intermediate positions. The way i write (here and in real life) is to overwrite and then prune later. There's no problem removing extraneous/unneeded info. But, mindless reduction according to some text > image style guideline doesn't necessarily make a good online encyclopedia. It's possible such a guideline is ill-conceived (and perhaps due to a historical print medium bias in which space does have significant costs).

I shrunk the diagram template by converting the larger pngs to svgs. That may help the page load time somewhat. (But, maybe it doesn't?) As you say, as the articles balloon up, they can be forked into smaller ones. For example, chess already splits tactics and strategy up into different articles. (This can probably be done for the aigakari stuff too -- if i keep at it, it's going to get at least 5 times longer.)

One can always ask other folks to comment on the amount of diagrams. Wikipedia is full of opinions.

The ideal solution is something like a video or a kifu-type thing. Unfortunately, videos are a pain to edit. (I'm definitely not motivated to do that.) I doubt wikipedia has any scriptable solution for a kifu reader. And, given that not many have done much with shogi strategy stuff in the past 8 years, I don't see this happening anytime soon. – ishwar  (speak) 21:59, 27 May 2016 (UTC)

Piece Values

I have no doubt that this YSS program is infinitely stronger than my (almost non-existing) Shogi abilities, but the values it assigns to promoted pieces as quoted here in the table just makes no sense. A promoted Silver is assigned a higher value than a Tokin. But it is much better to have a Tokin against the opponent having a promoted Silver than to swap the two around. Their tactical abilities are the same, as they both move as Gold, and the positions would only start to diverge once one of them gets captured. When my piece gets captured I would be off much better if it had been a Tokin, as my opponent then only gets a Pawn in hand, rather than a Silver...

I am not sure if the table gives a wrong interpretation of the values actually used in YSS, or if it is a genuine bug in the program. But should we propagate information we know to be wrong, without warning?H.G.Muller (talk) 06:39, 7 November 2018 (UTC)

P.S. By the same reasoning a promoted Silver should also be better than a Gold, as a Silver in hand will be less (for the opponent) than a Gold (although not much), while it is tabulated as being worth less. I conjecture that someone subtracted values that in reality had to be added, i.e. that a promoted Silver had to be Gold + 0.2, rather than Gold - 0.2, and that the value for a Tokin should thus have been 6.9 + 2.7 = 9.6 rather than 6.9 - 2.7 = 4.2. H.G.Muller (talk) 07:22, 7 November 2018 (UTC)

Well, i dont know much about YSS, but those are the values cited in a table in the paper referred to. (i checked again to make sure.) Whether these value truly make sense depends on the context of the rest of the program, which i can't evaluate since i dont know much about YSS. However, what you say does make sense: the values in YSS are different from professional players and other shogi engines in the way you mention. But, i can't say for sure this is propagating 'wrong' information. YSS is also a rather old program, and so probably not very strong (maybe it's because of these values, haha?).

I also note that the paper referred to actually cites a website for the YSS values. But, that website doesnt exist anymore, so i can't confirm the values in the paper or whether they calculated the value incorrectly (as you suggest above).

I guess the point of this section in the article is just to show that there is no agreement on specific values for shogi pieces as well as show that there is variation even on general piece value relations. In addition, to YSS's difference in the promoted values, YSS also shows a difference in adding an amount to the value when the piece is held in hand. I didnt find that in other lists of piece value and thought it was interesting to point out.

Do the YSS values in this article actually help the world understand shogi piece values? Well, i actually didnt really think that hard about it when i wrote this. My editing plan then was to dump alot of piece values that i found here and think about the big picture later. All this could be discussed more to give a kind of 'warning' about YSS as you suggest. – ishwar  (speak) 22:26, 8 November 2018 (UTC)