Wakawakawaka

That's an interesting idea. Here's what happens if you take the NORMSDIST of the player's score minus the AVG division score, divided by the std err of the top half of the results, and apply that to a top division percentile. Player - points - percentile - wins (using percentiles from completed basho 1959-present) First basho 199803 Ekigozan 35 0.967 13-2 Getayukata 34 0.946 12-3 Leonishiki 27 0.500 8-7 Imanonami 25 0.323 6-9 Okikuma 21 0.084 4-11 Bobokuni 20 0.054 4-11 Hypothetical Makuuchi from 202401 TimSumo 76 0.972 13-2 BlackPinkMawashi 76 0.972 13-2 Kaito 73 0.905 11-4 Heriokuno 73 0.905 11-4 TochiYESshin 72 0.866 10-5 Gurowake 72 0.866 10-5 Reonito 71 0.818 10-5 Shoryuyama 71 0.818 10-5 Gooner 70 0.76 9-6 Konosato 68 0.619 8-7 Golynohana 66 0.46 8-7 Goshobiyama 64 0.307 6-9 Asashosakari 64 0.307 6-9 Hidenotora 62 0.182 5-10 Cheesofuji 61 0.134 5-10 Chiyonosawa 60 0.095 4-11 Nantonoyama 60 0.095 4-11 Korinokoishi 59 0.065 4-11 Oskanohana 56 0.017 3-12 ScreechingOwl 56 0.017 3-12

Just submit your entry with your usual e-mail, but type in the new shikona. That's what I did anyways.

I love how enthusiastic the sumo community is with databases and statistics! I feel right at home!

@Gurowake No thoughts on the method I've outlined? It works pretty well. You can compute ranks formulatically also if i understand what you mean by that... say there are 10 players in juryo, j1e's mathematical rank is 2.0, j1w 2.1, j2e 2.2... j5w 2.9, ms1e 3.0. A 9-6 is worth +3 ranks in real juryo banzuke, in gtb banzuke it is worth +3/14 of the division or +0.2whatever EDIT: I also think a TD system might be useful (the player's debut basho would determine whether they qualify for a TD) Consider the following... 70 new players debuted in 2020. Of those that debuted with a 10-5 or better record (11); 4 reached the top ten at some point later on (with a 5th barely missing by reaching #11), and 3 were 'drop-outs', having yet to participate in 6 total basho. Of those that debuted 9-6 or less (70), 5 reached top ten at some point, and 44 were 'drop-outs'. So perhaps someone who gets a 10-5 or better in debut could be granted a MsTD, and someone who finished 8-7 or 9-6 could be granted a SdTD. The function of this would be 3-fold 1) Prevent Jk and Jd from getting too many 15-0's if we are to implement a division based scoring system 2) Accelerate players rank matching their ability. 3) From my experience trying this 'meta-GTB' exercise, I've noticed a lot of turnover of players. Many debut and participate in only a few basho before disappearing. This makes the meta-banzuke rather bottom-heavy, so a TD system would help keep the middle divisions full.

How about this? There are 6 strata (divisions). Take the average point score for each division, example. M = 53.4, J = 49.57, Ms = 52.89, Sd = 49.7, Jd = 41.77, Jk = 40.43 Rearrange those, so that the highest division has the highest value and the lowest division has the lowest value. M = 53.4, J = 52.89, Ms = 49.7, Sd = 49.57, Jd = 41.77, Jk = 40.43 Those would be the MK/KK lines for each division. Then comes the problem of amplitude, how many wins above or below the MK/KK line to give? Take the std. err. of something. The scores as a whole is usually around 9 but fluctuates between 6-12, but is heavily influenced by outliers at the bottom. Maybe apply some sort of outlier removal or take the std. err. of the middle 50% or something. Let's assume the std. err. is 9. If someone gets 9 points above the division average, that's 1 SD. Apply the NORMSDIST function in excel, that returns a value of 0.841 (think of it as a win %), multiply that by 15 'bouts', and you get 12.62. Round that to 13. Conclusion: 9 points above division average = 13-2, and consequently 9 points below division average = 2-13. Just as a fun exercise, using the above averages with a std. err. of 9, if an individual scored 54 points, it would be worth an 8-7 in M or J, 10-5 in Ms or Sd, 14-1 in Jd or Jk.

Bushozan now has 10 Juryo KK in a row, I'm not sure how to look it up, but that must put him pretty high on that list I'd assume ('greatest yo-yo rikishi of all time?' :P)

That happened last year too. It was all updated in one big chunk.

I believe I researched that question before, and IIRC the answer was there has never been 2 identical submissions in a game of GTB. I did get that right, and my logic was they wouldn't want to put Kagayaki right behind Chiyoshoma (the former being 1.5 ranks lower and with 1 fewer win than the latter) ____________ In regards to the GTB yusho... I'd like to thank my supporters, I tried not to think about the yusho, I just wanted to give it my all, taking it one guess at a time, and doing my brand of GTB. My goal now is to train hard and perform in a way that humbly honors the dignity of the top rank. :)

Just did a quick tally of Koto(nowaka/zakura)'s matches since becoming Sanyaku. My categories are surely highly subjective, but nevertheless... 'Pushers' (Onosho/Hokutofuji/etc.) 14-7 'Thrusters' (Daiesho/Abi/etc.) 26-8 'Yotsu' (Wakamotoharu/Hoshoryu/etc.) 20-21 'Chaos agents' (Midorifuji/Ura/etc.) 23-9 'Mixed style' (Takayasu/etc.) 6-1

He excels at being rather immobile. Pushers have a hard time moving him, and he doesn't move himself enough to overcommit and slip up against chaos agents.

Think Onosato is hurt or just trying something new? His style seems different than it has been the past 2 tournaments.

Terunofuji's 13-2 got him from M17 to M1e.

Whoever was doing commentary on Grand Sumo Highlights that day mentioned the connection. That's how I at least found out Terutsuyoshi retired.

He's the life of the party when featured on Dosukoi Sumo Salon, always smiling and joking! Looks like Ura paid some tribute to him today with Ashitori win :P

If we're going to include tribes from the Pacific Northwest, I believe we must include the Kwakwakaʼwakw as well.

Do people really like Wakamotoharu this much or do they just dislike whoever he's paired up against? I get he's a solid no-nonsense rikishi... maybe I just see him as too 'no-nonsense', i.e. dull. Also wouldn't it be totally wrong for Takayasu to lose this round and not be able to secure his JY?

Maybe Hakuho is so competitive and with such a big ego that he will persist out of spite, 'They think they can break the great Hakuho, eh? I'll show them! I will be 100x more humble than anyone has ever been!'

You know me too well.

Prefixes don't count. Then again, perhaps I should have been more strict and excluded acronyms for consistency. I will gambarize in my future foruming, and take it one post at a time.

Bob, Gog, lol, mom, pop, S.O.S., tot. Your list of 3 letter palindromes with middle letter 'O' is now complete.

It's not an egg though, it's some sort of mammalian epidermis (usually bovine?). Since it is also mammals handling the ovoid bovoid, the many referee discussions I'd assume are to help discern where the player ends and where the ball begins. Especially if the player ate a lot of beef before the game.

Very nice. So I think beyond adding an 8th item (a scoring system that addresses the 'everybody in Makuuchi getting KK' problem we discussed Tuesday) it sounds like you've got an efficient workflow figured out.

I don't really know anything about script, so to ask a different way, what parts do you still have to do manually?

What is the extent of the scripting/automation you have in place as far the entirety of the process goes? Like how long do you think it will take of human effort to create one of the newer banzuke with 200 players participating?

Just as a heads up, here are the shikona changes from the Jurassic until the end of 2003. Basho - Old Shikona > New Shikona 200011 Jasonishiki Spondinoumi 200011 Kitaseiya Kotoseiya 200011 Midorishima Tamanaogijima 200107 Wakatanuki Itachi 200109 Kurausu Kura-Uzo 200109 Magunoyama Maguroyama 200109 Itachi Wakatanuki 200109 Gebono Gernobono 200111 Kura-Uzo Kurausu 200201 Magicdidi Chiyotokimiya 200201 Kurausu Sakurai 200203 Purukumi Kaikitsune 200207 Sakurai Kurausu 200211 Makususuchi Zenjimoto 200307 Feginowaka Inazuma 200309 Wakanozan Gibmaatsuki