Don’t want to wait for the official banzuke announcement on August 27th? The Crystal Ball is here to give you a good idea of how it’s likely to play out.

Upper San’yaku

Y1	Kakuryu	Hakuho
Y2	Kisenosato
O1	Goeido	Takayasu
O2		Tochinoshin

No changes here following the results in Nagoya. Tochinoshin will be kadoban at Aki after his withdrawal, and will need 8 wins to maintain his Ozeki rank.

Lower San’yaku

S	Ichinojo	Mitakeumi
K	Tamawashi	Takakeisho

Takakeisho was the only rikishi from the upper maegashira ranks to earn double-digit wins, making him the obvious choice to take over the Komusubi slot vacated by Shohozan. While it seems obvious that yusho-winning Mitakeumi (13-2) should be ranked ahead of frequently lethargic Ichinojo (8-7), the current banzuke committee seems to strongly favor recent precedent over common sense, and so I’m predicting that they won’t switch them from their current ranks. For once, I’d be happy to be wrong.

Rank-and-file (Step 1)

I thought I’d do something a little different for this prediction post. How do I make these forecasts? It’s basically a two-step exercise. Step 1 is entirely algorithmic: where “should” each rikishi in Makuuchi be ranked for the next tournament, given their current rank and their performance in the just-completed basho? This produces a list like this:

M1:   Yutakayama, Ikioi, Kaisei

M2:   none

M3:   Chiyotairyu

M4:   Shodai

M5:   Chiyonokuni, Endo

M6:   Abi, Myogiryu, Onosho, Asanoyama

M7:   Kagayaki

M8:   Takarafuji, Tochiozan

M9:   Shohozan, Hokutofuji

M10: Kotoshogiku, Daishomaru, Daieisho, Aoiyama

M11: Sadanoumi

M12: none

M13: Kyokutaisei, Nishikigi, Okinoumi

M14: Ryuden

M15: Chiyoshoma, Chiyomaru

M16: Yoshikaze, Ishiura

Juryo: Arawashi, Meisei, Kotoeko

Looking at this list reveals some obvious problems. Some ranks have too many rikishi (e.g. M6 and M10, with four each), while others have too few (or none, in the case of M2 and M12). Since every rank must contain exactly two rikishi on the actual banzuke, this is where “banzuke luck” comes in: some rikishi will be ranked higher and others lower than they “deserve.” Most of the time, rikishi from (say) the M5 group will be ranked above those from the M6 group, but the decisions for whom to move up or down from within a group to fill out the ranks are much more subjective. Note also that I’ve left off the list my predicted promotions from Juryo (Takanoiwa, Takanosho and Kotoyuki), as these computed ranks don’t translate well between divisions.

Rank-and-file (Step 2)

In step 2, I go through the list above and decide which rikishi to move up/down so that each rank is appropriately filled. I also have to decide whom to place on the East/West side, and where to slot in the guys from Juryo. Some decisions follow simple rules or are otherwise obvious, while others are basically some combinations of gut reaction, coin flip, and (often fruitless) attempts to guess what the banzuke committee is thinking; the latter is made difficult by the fact that they often make inconsistent decisions in similar situations from one banzuke to the next, or even within a single banzuke.

I won’t go through the whole thing, but let’s walk through a few examples (or skip to the next section to see the predictions). Which of Yutakayama, Ikioi and Kaisei should move down to M2? Well, M2 is already a huge promotion for M9 Yutakayama, and the banzuke committee doesn’t like really big jumps up the banzuke (see Chiyonokuni, last basho), so Yutakayama is the odd man out, moving down to M2e. He’s joined at M2w by the next highest-ranked rikishi, Chiyotairyu. Shodai claims one of the M3 slots, and I favor Endo over Chiyonokuni for the other, given Endo’s popularity and winning record and Chiyonokuni’s withdrawal and two fusen wins.

An example of a tricky area is how to sort out the four rikishi who “should” be ranked M6. The numbers dictate that one of them will move up to M4, two to M5, and one will be ranked at M6. But how to order them? By previous rank? By win total? By KK vs. MK? There’s no obvious answer, and no consistent pattern to what the banzuke committee has done in the past.

As we continue down the banzuke, the gaps at M11, M12 and M14 create natural openings for Takanoiwa, Takanosho and Kotoyuki, respectively, and we get down to the final four slots, which seem relatively clear-cut.

The predicted maegashira banzuke

And so we arrive at the final product. I hope the process description gives the reader a better sense of how I got here, and which of these placements are more solid or wobbly.

M1	Ikioi	Kaisei
M2	Yutakayama	Chiyotairyu
M3	Shodai	Endo
M4	Chiyonokuni	Onosho
M5	Asanoyama	Myogiryu
M6	Abi	Kagayaki
M7	Tochiozan	Shohozan
M8	Takarafuji	Hokutofuji
M9	Kotoshogiku	Daishomaru
M10	Daieisho	Aoiyama
M11	Sadanoumi	Takanoiwa
M12	Takanosho	Kyokutaisei
M13	Nishikigi	Okinoumi
M14	Ryuden	Kotoyuki
M15	Chiyoshoma	Chiyomaru
M16	Yoshikaze	Ishiura