Sumo Reference Updates

[Asashosakari 18,786]

I just noticed a very light error in the database on the kimarite list for ozeki Takanonami in the Kyushu basho 1997.

http://sumodb.sumogames.com/Rikishi_basho....=5&b=199711

Actually, Takanonami won the playoff bout against Takanohana by Kawazugake (What a beautiful bout!), and not by uwatenage. I don't remember his bout against Tochiazuma on senshuraku, but I don't think the kimarite was again the very rare kawazugake.

Hmm, sorry, but I rather stick with my black-on-white source, the Sumo magazine - which might be wrong or not. But kimarite are often decided differently from what you might feel. So I certainly will not correct this "error", unless someone presents me with a better source.

The kawazugake playoff win was Takanonami's first yusho in 1996.01, not the second one in 1997.11 - Chienoshima, see here for an SML post from Kyushu 1997 which makes it clear that the bout in the Youtube video you linked on the SML yesterday can't be the one from 1997.

Edit: Hmm. (I am not worthy...)

The great Takanonami beat Takanohana by kawazugake in a playoff bout to clinch his second and final yusho in november 1997. This is the most unusual and possibly the most beautiful bout I have ever seen : Takanonami style !!!! I wrote an article on this bout for the 4th number of Le Monde du sumo (english translation here)

With the help of excellent footwork, all the while going backwards, the Ozeki wraps up Takanohana's left leg, thus depriving him of precious support, and topples him roughly backwards, with his right arm which relaxed its grip in order to get hold of the knot of Takanohana's mawashi at his back. Kawazugake! Great artistry in the space of 23.4 seconds, it is a pure wonder of a bout, which will remain for all time in the bouts of folklore. It is an unusual bout by Takanonami, as much in conception as in execution, and a well-earned first yusho.

Edited July 16, 2008 by Asashosakari

[Chienoshima 0]

I just noticed a very light error in the database on the kimarite list for ozeki Takanonami in the Kyushu basho 1997.

http://sumodb.sumogames.com/Rikishi_basho....=5&b=199711

Actually, Takanonami won the playoff bout against Takanohana by Kawazugake (What a beautiful bout!), and not by uwatenage. I don't remember his bout against Tochiazuma on senshuraku, but I don't think the kimarite was again the very rare kawazugake.

Hmm, sorry, but I rather stick with my black-on-white source, the Sumo magazine - which might be wrong or not. But kimarite are often decided differently from what you might feel. So I certainly will not correct this "error", unless someone presents me with a better source.

The kawazugake playoff win was Takanonami's first yusho in 1996.01, not the second one in 1997.11 - Chienoshima, see here for an SML post from Kyushu 1997 which makes it clear that the bout in the Youtube video you linked on the SML yesterday can't be the one from 1997.

(I am not worthy...) (In jonokuchi...)

Yes, you are both right, it was the first yusho in 1996.01 !! I should have checked your database before sending my mistake twice !!

Excellent database anyway ! I did not know Takanonami was such a Kawazugake master !! I miss him !

Greetings from Paris !

[ryafuji 807]

I assume this one is a glitch, as I haven't come across anything like it before, but Chiyohakuho is recorded as dropping out of the March 2005 basho without losing a bout by default:

http://sumodb.sumogames.com/Rikishi_basho....98&b=200503

[Asashosakari 18,786]

I assume this one is a glitch, as I haven't come across anything like it before, but Chiyohakuho is recorded as dropping out of the March 2005 basho without losing a bout by default:

No, that's correct.

(But, how on Earth did I produce that list at the bottom of the thread?)

[ryafuji 807]

I assume this one is a glitch, as I haven't come across anything like it before, but Chiyohakuho is recorded as dropping out of the March 2005 basho without losing a bout by default:

No, that's correct.

(But, how on Earth did I produce that list at the bottom of the thread?)

Tsk. How could I have doubted the database. Interesting that it was by no means unprecedented either.

[Sakana 46]

Doitsuyama, I sent you a picture of Nagakura (~jonokuchi 37) some weeks ago but I don't know if it was ok or not... (sorry for lil' off-topic)

[Doitsuyama 1,173]

Doitsuyama, I sent you a picture of Nagakura (~jonokuchi 37) some weeks ago but I don't know if it was ok or not... (sorry for lil' off-topic)

Ah yes, I remember, I didn't come to that yet. Actually, rikishi pictures are in the hands of Fay who sends them to me for uploading.

[Sakana 46]

Ok, thanks for the reply. (Sign of approval...)

[Doitsuyama 1,173]

It's been a while but there's an update to Sumo Reference: I have added all full banzuke from 1934.05 to 1966.09 which makes it 115 full banzuke more. Now there are 368 full banzuke in the database (from 1934.05 to now), leaving about 350 partial banzuke (from 1757 to 1934.01). This will likely stay so since I have no sources for full banzuke prior to 1934.05.

In fact, I want to thank Gans-san for providing me with hoshitori-hyo scans for all banzuke from 1934.05 to 1988.11. Without them I wouldn't have been able to make the database as complete as it is now.

[Randomitsuki 2,768]

I have added all full banzuke from 1934.05 to 1966.09 which makes it 115 full banzuke more. Now there are 368 full banzuke in the database (from 1934.05 to now)

Thank you very much for an accomplishment of epic proportions! :-O (Applauding...)

[Doitsuyama 1,173]

I agree that the banzuke and bout queries are a bother to copy when there are many results, but a fairly simple solution would be a way to allow more results per page to display.

Heh, I faintly recall asking for that as a backdoor option quite a while ago. Of course, that was back when the DB still regularly killed the performance of the sumogames server, so perhaps it wasn't entirely advised at the time. If it's back in play I'd definitely appreciate an option to have something like 500 or 1,000 lines per page.

But while we're making requests - two others that I'd like to reiterate from long ago, for the banzuke query: 1) Sortability of grouped results like the bout query already has, and 2) the ability to add more (fixed) data to the output table such as birthdates and shusshin. I'd love to be able to do quick Excel work such as calculating how old each rikishi was at [division] debut or how many juryo there have been from [shusshin], but right now it's troublesome to have to pull that data from elsewhere, where elsewhere usually means "the full Rikishi.aspx output".

You will be delighted to see that this wishlist is done. At the top there is a new combo box with an option of rows per page. Further down is a new section with optional rikishi details to include in the output, and you can sort grouped result (also by the new rikishi details).

I realize that this enables even more wishes like new search conditions (all sekiwake heavier than 160 kg and the like) or new group options (for example group by heya), but this will have to wait.

[Asashosakari 18,786]

You will be delighted to see that this wishlist is done.

Delighted I am! (I am not worthy...) (I am not worthy...) (I am not worthy...)

[Doitsuyama 1,173]

Also, if I could only figure out how to spit out the last 6 basho's results in the lower divisions when the young'uns haven't had that many...

Ok, I think I found a nice solution to your problem discussed in some detail the original thread. The base basho for the query (which used to always be the first basho) now is the first basho with a condition. So, if the only condition is in the sixth basho, that's the base basho for the query, and the first five get added after that. Now no rikishi from the first basho with a condition is going amiss which should be sufficient I think. This required some reprogramming so some things might go awry because of that - I already detected and fixed a few issues.

[Doitsuyama 1,173]

It's been a while again since the last major update, but now I'm elated to announce something new. Tamanaogijima and Asashosakari put in the amazing work to get a detailed kabu history since 1927, and it's now included in Sumo Reference: http://sumodb.sumogames.com/Kabu.aspx. I think it's very useful since detailed information about kabu is somewhat hard to get, even in Japanese.

You can browse through the history with the links in the tables (the rank of a holder or owner links to his kabu history, the shikona links to his rikishi profile), or select the history of a specific kabu or rikishi with the combo boxes on the left. You can also get a complete list of kabu for any given date since 1927 - of course such a list might be wrong for a few kabu since we often know only the month of a change, not the exact date.

If you detect errors or can improve on the details, please contact me, so I can make the corrections. Also, the Japanese translation is quite rough (or even non-existent in the changes legend), so if someone could go over the Japanese side and improve it, I'd be happy too.

[Fay 1,677]

Great!

Thanks a lot for the work!

[Asashosakari 18,786]

In fairness, it was more like

Tamanaogijima and Asashosakari

(Holiday feeling...)

[kame 0]

(Holiday feeling...) Amazing work. Thanks to everyone involved.

[Gusoyama 95]

amazing stuff as always.

[shumitto 418]

Icing on the cake...

[Oshirokita 173]

oooooh, awesome!

I am just salivating at the prospect of diving into the kabu database.

Thanks guys!

[Manekineko 200]

Just to add my (Showing respect...) to all those working on the Doitsubase. I've been writing short basho reports for few sumo-deprived Croatians since Eurosport stopped airing, and the Reference is a huge help. It's so amazing we have this. Utterly. (Holiday feeling...)

[Asashosakari 18,786]

Ok, I think I found a nice solution to your problem discussed in some detail the original thread. The base basho for the query (which used to always be the first basho) now is the first basho with a condition. So, if the only condition is in the sixth basho, that's the base basho for the query, and the first five get added after that. Now no rikishi from the first basho with a condition is going amiss which should be sufficient I think. This required some reprogramming so some things might go awry because of that - I already detected and fixed a few issues.

Seems I've managed to break it somehow...this query fails to include the 19 current jonidan rikishi who only started their career in 2009.03. Not sure what's wrong there as I've seen the above-mentioned fix work just fine in other queries...

Hmm, upon checking: Seems that this happens when there are unused query fields (with lower basho numbers) in front of the one that actually includes restrictions. Compare this and this. (Difference only visible in show_form=1 mode.)

And another sort-of-related issue - as seen in the query above, rikishi with empty basho fields get no name. Those lines do correspond to actual rikishi, but (besides the missing names) the Rikishi.aspx links are all busted. Looks like an overflow issue... Can also be triggered by making reference to basho with not-yet-existing data. Examples: Query set on basho=2 and year=200907, another empty field also references basho=2 ---> displaying too many columns breaks it (compare this and this); query set on basho=2 and year=200907, another empty field references basho=3 or higher ----> display always broken (here). This bug variant doesn't seem to depend on the order in which the query fields are used (i.e. form1_basho_nr=2 and form2_basho_nr=3 breaks it as well as form1_basho_nr=3 and form2_basho_nr=2.)

[Doitsuyama 1,173]

Ok, I think I found a nice solution to your problem discussed in some detail the original thread. The base basho for the query (which used to always be the first basho) now is the first basho with a condition. So, if the only condition is in the sixth basho, that's the base basho for the query, and the first five get added after that. Now no rikishi from the first basho with a condition is going amiss which should be sufficient I think. This required some reprogramming so some things might go awry because of that - I already detected and fixed a few issues.

Seems I've managed to break it somehow...this query fails to include the 19 current jonidan rikishi who only started their career in 2009.03. Not sure what's wrong there as I've seen the above-mentioned fix work just fine in other queries...

Hmm, upon checking: Seems that this happens when there are unused query fields (with lower basho numbers) in front of the one that actually includes restrictions. Compare this and this. (Difference only visible in show_form=1 mode.)

You are right, I didn't check if query fields were used or not. I enhanced that, and unused fields are ignored now. In the process I made some improvement on the query URLs as all parameters with default values are removed which makes for much shorter links.

And another sort-of-related issue - as seen in the query above, rikishi with empty basho fields get no name. Those lines do correspond to actual rikishi, but (besides the missing names) the Rikishi.aspx links are all busted. Looks like an overflow issue... Can also be triggered by making reference to basho with not-yet-existing data. Examples: Query set on basho=2 and year=200907, another empty field also references basho=2 ---> displaying too many columns breaks it (compare this and this); query set on basho=2 and year=200907, another empty field references basho=3 or higher ----> display always broken (here). This bug variant doesn't seem to depend on the order in which the query fields are used (i.e. form1_basho_nr=2 and form2_basho_nr=3 breaks it as well as form1_basho_nr=3 and form2_basho_nr=2.)

This was some leftover issue from changing the "database start basho" to the first used basho. I corrected that too.

[Asashosakari 18,786]

Did the data entry fields get removed from Rikishi.aspx when it's in individual profile mode?

[Doitsuyama 1,173]

Did the data entry fields get removed from Rikishi.aspx when it's in individual profile mode?

Yep. The pages load much faster this way.