By popular demand, let’s take a look at the July banzuke and how the Crystal Ball fared (spoiler: not very well). It started off strongly, getting the first 11 slots, covering the named ranks and M1, exactly right. This wasn’t exactly challenging though—most of these ranks were set in stone. At the ones that weren’t, Terunofuji’s ranking as the top Ozeki, leapfrogging the previously higher-ranked Takakeisho, confirmed the primacy of a head-to-head playoff win. Meisei, making his san’yaku debut, got the nod over Endo for the West Komusubi rank, and the incumbent Daieisho only dropped one rank to M1w after posting a 6-9 record.
I was mildly surprised to see Takanosho, 5-10 at S1w, ranked just ahead of Ichinojo (9-6 at M6w), but both men are at M2, as predicted. And the streak of correct predictions ends with Hokutofuji, who was a lock for M3e. The biggest shock came at the next position, M3w, where we find Tobizaru dropping only a single rank after going 5-10. A demotion of this leniency is virtually unprecedented. I looked at the hundreds of cases of 5-10 records in the M1-M5 ranks during the 6-basho era, and could find only a single comparable instance, back in 1967, when someone was demoted from M3e to M4e. Moreover, the banzuke committee was by no means forced into this choice—they had to over-promote Kotoeko and either Okinoumi or Chiyotairyu anyway, and it would have been much more palatable to place them ahead of Tobizaru.
The next surprise to me was the decision to dramatically over-promote all three of Kotoeko, Okinoumi, and Chiyotairyu, when only two of these were forced. I get the point about trying to treat them similarly, but none of them even remotely deserved to be ahead of Hoshoryu and Onosho, who should have retained their ranks after posting 7-8 records, as has been the custom recently and as happened lower down on this banzuke with Tamawashi and Terutsuyoshi. In a similar vein, why in the world did they feel compelled to promote Chiyoshoma 4 full ranks after an 8-7 performance, pushing down 7-8 Takarafuji in the process?
The Crystal Ball also missed Chiyomaru’s ranking by a whopping 2.5 ranks, although this is probably my fault for placing a kachi-koshi top-division rikishi below the promotions from Juryo, even if Ura deserved to be promoted 5 full ranks ahead of Chiyomaru “by the numbers.” And in a final surprise, I had Yutakayama claiming the last promotion slot instead of Ichiyamamoto. Getting to Makuuchi with an 8-7 record from J4 is rare, but not as rare as doing so from J8 with a 10-5, and on top of this, Yutakayama should have gotten the benefit of the doubt by virtue of his considerable top-division experience.
Overall, the Crystal Ball got only 17 of the 42 ranks exactly right, and placed an additional 11 rikishi at the correct rank but on the wrong side—one of my worst performances of the last two years. Of the 14 misses, 7 were by half a rank (if we count M17e/J1e), 4 by one rank, and 3 by more than a rank—the aforementioned Chiyotairyu and Chiyomaru, as well as Tokushoryu. Hopefully, some of this analysis will improve future banzuke predictions, but some of the misses can only be chalked up to the banzuke committee departing from both historical precedent and internal consistency. I often think how different things might be if they had to publicly justify their decisions, as is the case in most other contemporary sports.