
The Kyushu banzuke has been released. How did my prediction do?
The Good
The prediction got the named ranks exactly right, as well as the M1 rank, for 11 perfect predictions at the top. Admittedly, the only real question mark here was whether Fujinokawa’s 11 wins would be enough for him to leapfrog Atamifuji, which he did. The prediction was also perfect from M13 to M17, getting all the exchanges between Juryo and Makuuchi right, as well as the placement of those being promoted. In all, this is 20 ranks nailed exactly at the top and the bottom of the banzuke.
The Bad
Maybe calling it “bad” is a little harsh, but from M7 to M12, the prediction placed only 6 of the 12 rikishi at the exact position, plus 3 more at the correct rank but on the wrong side. All of the misses were the result of just two puzzling decisions by the banzuke committee. M4e Takanosho (4-9-2) got zero benefit for being ranked in the joi and fighting all the top guys, dropping a full 7 ranks, below Asakoryu, Shishi and Oho, who all got over-promoted or under-demoted as a result. But then, M6w Ichiyamamoto, who likewise had 4 wins and a much easier fight card, got a soft 5.5-rank demotion, placing ahead of Abi despite deserving to be a full rank lower by the numbers.
The Ugly
The prediction really went off the rails from M2 to M6. It had only one of the 10 rikishi at the correct rank, and even he ended up on the other side. Lots to complain about here. Why did M2e Takayasu (7-8) not hold rank, when he faced the toughest possible schedule of opponents and deserved to be ranked ahead of Ura and Fujiseiun by the numbers? Why did the committee feel the need to squeeze Takerufuji between the two demoted komusubi, dropping Daieisho to M4w? Why did Kotoshoho, Kotoieho, and Gonoyama get zero credit for their joi schedules?
We’ll have to see if the lack of respect for higher ranks was an idiosyncrasy of this banzuke, or if it’s a new trend. Now, it’s on to the basho!
Discover more from Tachiai (立合い)
Subscribe to get the latest posts sent to your email.
I too was puzzled by the overpromotion of Ura, Takayasu’s drop and Daieisho being put lower than Hakunofuji. Identical records at the same rank should have had them together at M3 (where I put them) or M4.
The past practice of 7-8’s staying put also seems to have gone away.
Just from personal experience in various group talks, i’d suspect that there are more deals between banzuke designing oyakata plus regarding ichimon connections as we could possibly discover only in the top division. Like, ‘if we agree to push up your Makushita promotee/post injured returnee from Jonidan/sinking star in Juryo, you have to give in with somebody else of your stable’. Who knows? Maybe in near future you could tell an AI analyst to pull a string out of these and get on trail where it’s ending at…
I‘m suspecting that there are AI guesses right now already. I‘d like to know how they did.
When I’ve tried, they’re terrible, but I haven’t really played with the most advanced models
The forte of AI is to find patterns in a lake of data. But of course there is no such lake with banzuke data. The 154 (?) rankings of the century do not even build a puddle. Therefore even the best versions of AI might not have a great advantage compared to U.
I recently googled the Takayasu vs Onosato bout from the Nagoya 2025 basho and the AI told me that Takayasu *won* that basho, so I’m not holding my breath ;-)
Ain’t nobody got time for that kind of haggling, they’re assigning 600-ish ranks in only a few hours. They clearly don’t even have time for quality control beyond the absolute basics. (Check out where Fukutsuumi and Mihamaumi landed in jonidan…)
I had 32/42 exact guesses and 2 same row guesses for 66 points.
My guesses were off as follows:
for M2e-M3E I had Takayasu, Ura, Fujiseiun (similar to discussion above);
For M4E-M6W I had Daieisho, Kotoshoho, Kotoeiho, Gonoyama, Takerufuji, Hiradoumi (similar discussion as above).
For M12W to M13E I had Toshinofuji, Abi, Ichiyamamoto; Actually I had Abi then Toshinofuji in my spreadsheet but made an error on entry. So I got 2 points for being incompetent.
Is there any word on Wakatakage’s recovery and progress? Will he be able to compete in November pool? If so, how far has he fallen?
While Wakatakakage has not yet resumed bout practice, he has restarted collision drills (butsukari-geiko), stating, “I’ll take things without rushing.” He focused on basic physical exercises at Arashio stable while attending rehabilitation sessions three times a week. His body weight, which dropped after the surgery, has returned to the upper 130-kilogram range, matching his peak condition. Ranked at West Juryo 5 for the upcoming Kyushu tournament, a full absence would inevitably drop him to the makushita division for the January 2027 tournament.
(Sanspo)