What Does it Take to Become Ozeki?

Takakeisho, the most recent addition to the rank of champions

As a break from the news, let’s revisit what it takes to reach sumo’s second-highest rank. With two aging Yokozuna and only one active Ozeki, we certainly need reinforcements. I took a long look at Ozeki promotions in the six-basho era about a year ago, and would encourage you to look at that post. Since the standards have evolved over time, here I’ll restrict my attention to the past 30 years, during which time there have been 25 Ozeki promotions.

A commonly held belief, echoed by Wikipedia, is that reaching Ozeki takes 33 wins over 3 tournaments while ranked in san’yaku:

A wrestler at the rank of sekiwake will be considered for promotion if he has achieved a total of at least 33 wins over the three most recent tournaments, including ten or more wins in the tournament just completed. Promotion is discretionary and there are no hard-and-fast rules, though a three-tournament record of 33 wins is considered a near-guarantee.


Here’s a sumodb search that roughly matches these criteria: first basho ranked at least M4, second kosumubi or sekiwake, third sekiwake, with winning records in all 3, at least 10 wins in the third, and 31+ wins over the 3 tournaments. It returns 44 results—the aforementioned 25 promotions and 19 non-promotions. The latter include all 9 instances in which the win total is only 31 (most recently, Asanoyama’s last 3 basho), which I added as a sanity check. While promotions with as few as 28 (!!!) wins happened in the past, there haven’t been any with fewer than 32 since 1985.

So how about the apparent requirement for 33 wins, and the “near-guarantee” they supposedly confer? 35+ wins have indeed been a historical guarantee, with 10 promotions out of 10 instances. Rikishi with 33 and even 34 wins, however, haven’t been locks for promotion: there were only 6 promotions out of 9 for those with 34 wins during their “run”, and 6 out of 8 for those with 33, for a total of 12 out of 17 in this “likely but not certain” range. Takakeisho, of course, famously missed out on promotion with 33 over 3 a year ago, before earning it with 34 over 3 the very next basho.

And just as 33-34 wins isn’t a guarantee, 33 isn’t a requirement—we’ve seen 3 promotions out of 8 for those with 32 wins. This isn’t the 70% rate for those with 33-34, but at 38%, it’s not exactly a rare exception. Both Goeido and Kisenosato were promoted with 32. These numbers will be interesting to keep in mind going forward, as Asanoyama’s total currently stands at 21 over 2 basho, which suggests that he could be promoted after the next basho with as few as 11 wins, but needs 14 to be a lock.

26 thoughts on “What Does it Take to Become Ozeki?

  1. Some mention should be given to the “while in san-yaku” part, which has been overlooked for Terunofuji – if indeed the shimpan department even counted 3 basho at all, as that 9-win basho was a lukewarm achievement and out of san-yaku.

    • And of course for Tochinoshin, who started at M3, although he did accumulate 37 over 3, tied for the most ever. Those two are the only ones since 1990, but there are other examples earlier. That puts it at 2 out of 4 for those with 33+ wins and first basho as upper maegashira.

      • If we restrict the first basho to san’yaku, it’s 6 out of 7 with 34, 5 out of 7 with 33, and 3 out of 6 with 32 wins.

    • …and of course, he’s finished runner-up in the last two basho, which is a vanishingly rare feat for a maegashira.

    • The closest parallels are Takanosato M12 12-3 J M1 13-2 J S 11-4 (36 over 3), not promoted and Kotooshu M5 10-5 K 12-3 J S 13-2 D (35 over 3), not promoted.

      • In Kotooshu’s case there were 3 active ozeki and a yokozuna named Asashoryu who didn’t seem to be retiring anytime soon. In Takanosato’s there were 2 ozeki and 4 yokozuna (as well as an upwardly mobile Chilyonofuji at S1e). No-one was desperate for a new ozeki then, but they are now. I think Shodai will be promoted with a yusho and will have a serious chance with a 13-2 j-y.

          • Shodai getting promoted with a yusho would be not that far off from how Terunofuji did it, counting the third-last basho as worthless in either case. Sure, M4->S instead of S->S, but championships have had a tendency to make otherwise borderline promotion cases into no-brainer decisions. (Dejima and Chiyotaikai come to mind.)

  2. For those of you who get the TV Japan cable channel, this coming Monday evening’s episode of ‘Tsurube’s Salute to Families’ will feature Ajigawa Oyakata (the former Aminishiki).

    If you’ve never before seen this NHK show, in each episode actor/rakugo comedian Tsurube and a celebrity guest drop into a small town in Japan, then just walk around the town, meeting people on the streets and barging into people’s homes and workplaces, trying to get a sense of the hopes, dreams, and lives of small-town folks.

    It should be entertaining to see how people react to the beloved former rikishi.

  3. What happens if controlling for number of Yokozuna and Ozeki? Presuming it is harder to achieve 10 wins with more opponents of these ranks, I would expect fewer wins needed with more existing sanyaku…though it does seem counter intuitive. But think of it this way, there are more chances for “quality” wins that may help promotion chances and fewer bouts against maegashira.

    • Sometimes I wonder if it’s not a more important quality in an Ozeki to be able to be Maegashira and Komusubi consistently, thereby making the yusho race more interesting.

      • Yes. Somebody like Chiyotaikai got a lot more respect for beating up on the lower-rankers for 7-3 or 8-2 starts, even if he might have struggled against the high-rankers later on to only finish 10-5-ish a lot, than somebody like Kotooshu did for his Houdini acts in salvaging his tournament records against ozeki and yokozuna after 5-5 starts or worse.

    • I think it is actually the other way around.
      When there are ‘too few’ Ozeki/Yokozuna (remember this used to be just one rank) they look for someone to quickly fill the position. While when there are lots of men already holding the rank, they get quite picky.

      • Look at the banzuke from 2000 to 2002. It was an ozeki clown car back then, and they were not at all shy of promotion.

        • That was another period where the long-time high-rankers were falling to pieces though, and it came on the heels of a 5-year period without any ozeki promotions at all. That’s exactly when we would expect the promotion standards to be relaxed. It just so happened that a whole bunch of rikishi ended up fulfilling those lax standards nearly at the same time. It would have been a rather hard sell to the public that records that were good enough to see one guy promoted were no longer good enough for somebody else only a couple of basho later just because the ozeki rank now had two more people in it.

  4. This just in, the Sumo Association just finished their hours long deliberation and have decided the Osaka basho will commence however sans live audience!

  5. I think there is a requirement to always have 2 O, but since Y can step in as O, it essentially means that the sum of Y and O must be at least 2.

    That said, I wonder what they would do if that number drops to 1. We could soon reach that if all the promotion candidates squander their run, and then Hak & Kak retire.
    Promote someone with 31 or even less? Or take someone who does not have 3 KK in a row but still a good number of wins (like 7-13-13)?

    • I think they’d just ignore the “requirement” if there were no viable promotion candidate(s).

  6. Perhaps I have overlooked it, but I wonder how have rikishi performed historically, who were promoted to ozeki easily (without rigorously obeying to hardest standards), compared to those who were overlooked (at least firstly) allthough they got 33 wins. A sort of Kisenosatos against Takakeishos. I think that is a way to assess the quality of the people responsible for promotions to ozeki

      • Yes, thankyou – I have read that, very interesting work, but unless I overread it, it doesn’t compare the extremes. You compare to the average. Kisenosato is an example that relaxing the criteria did not produce the Japanese yokozuna all had hoped for (Bad luck because of his injury) Takakeisho could be an example also speaking against the wisdom of the promotion guys (nobody knows – the future will tell) Statistically all the bad luck should average out:
        I tried to figure it out (Correct my errors plesse)
        Kisenosatos: 56% Yokozunas, on average 4,8 yushos
        Takakeishos: 83% Ozekis ,21% Yokozunas, on average 3.9 yushos
        So the Kisenosatos still win, but not by much


This site uses Akismet to reduce spam. Learn how your comment data is processed.