Last week I looked at the 2024 yusho year and tried to situate it historically through focusing on the wins and rank of the yusho winner. I concluded by observing 2024 continued a trend beginning in 2019 of below-average win totals and a below-average rank of the champion. Some responded that I was being negative about the quality of sumo. Similarly, some said the metric is only useful for making assertions about dominance, not quality.
What I commented under that post, I repeat here: I was not attempting to be negative about the current state of sumo. Anything I write about Makuuchi division sumo is written with the understanding I am writing about the 40 or so best athletes in their sport. Every other month we are treated to the 40 best men in the world competing against one another. What other sport can say that? We are spoiled.
As for the critique of dominance versus quality: it is accurate…to a point. I don’t think it is unreasonable to assert a Yokozuna is a qualitatively better rikishi than a komusubi. I feel pretty confident that 14 wins is qualitatively better than 12 wins. If I had the choice between watching a year 6 komusubi yusho with 12 wins or 6 yokozuna yusho with 14 wins; I’d take the latter. In 2024, we basically saw the former.
So is the current state of sumo so bleak? No.
A shortcoming of the championship quality metric is the built-in dominance factor: it does not consider anyone else but the champion. Is there a way to evaluate the competitiveness of a basho, or an entire year of basho?
A simple way would be to compare the difference between the amount of wins between the yusho winner and the runner up. We could call this yusho differential: yusho wins minus jun-yusho wins=yusho differential. A small yusho differential (YD) would seem to indicate a more competitive yusho since it would mean the winner and runner-up had a similar number of wins. I am arguing that a 13-3 (Y)-12-3(JY) is more competitive than a 15-0(Y)-12-3(JY).
In a yusho year we would expect a YD close to 6 since the champion usually has at least one more win than the runner up. In fact, plotting the YD since 1958 this is what we see:
Most years, the YD ends up at 8 (12 times), closely followed by 9 (10 times) and 7 (9 times). So, a YD of 7 to 9 is average and indicates the champion winning 1 or 2 more matches than the runner up. Anything below this 7-9 range would indicate a more competitive basho and a higher total would indicate a less competitive basho.
Evaluating basho years this way, 1997 was clearly the most competitive ever with a YD of 2. Playoffs decided 4 of the basho that year: highlighted by March when there was a 4-way playoff! At the other end of the spectrum, 1984 and 2013 saw the champion outpace the runner up by 15 wins. 2013 may have been exciting for fans of Hakuho and Harumafuji, but probably not so much for others.
How did 2024 stack up in this measure of competitiveness? With the help of 2 playoffs, the YD for 2024 was 5. And believe it or not, this year was the third consecutive year with a YD of 5. We have had 3 consecutive years of basho with multiple playoffs (two in ’22 and ‘24; and three in ‘23). The YD metric indicates we are in a very competitive- some might even say, exciting- era of sumo! The past three years have been very competitive with yusho frequently coming down to the final matches of day 15. The only span of 3+ years that can equal this level of competitiveness was 1997-2001:
| Year | YD |
| 1997 | 2 |
| 1998 | 8 |
| 1999 | 4 |
| 2000 | 6 |
| 2001 | 4 |
Numbers don’t tell the whole story. Numbers will never tell us which Shodai is going to show up any given Tuesday. But numbers can help us tell a better story. Considering yusho-differential helps us appreciate the competitiveness of the past several years- one of the most competitive time periods in recent history. The lack of dominant yokozuna/ozeki performances have made way for closer yusho races. What excitement will 2025 bring?
Discover more from Tachiai (立合い)
Subscribe to get the latest posts sent to your email.
Perhaps another factor in calculating the competitive scale could be the number of rikishi that ended each basho with a Jun-Yusho. Having several wrestlers competing for the cup at the end makes for more interesting tournaments. the 5-way playoff at Kyushu 1996 and the 4-way at Haru 1997 had to have had huge interest.
One critique that I have is that this is still focused on the yusho. Of the 42-man roster, only a couple of guys will be “in the running”. These metrics are thus really only limited to the competitiveness of the top of the top. Realistically, it’s a smaller group of wrestlers than the sanyaku in any given tournament. The yusho race is also something that only really comes to the fore when the tournament is in its final days. There’s a lot more action and a lot more quality throughout the tournament, up-and-down the banzuke but they seem irrelevant to these metrics.
Yes, that’s right.
Competitiveness presumably means how much lower-ranked rikishi have a chance to beat higher-ranked rikishi.
If so, one way to measure competitiveness might be to just give a score to every bout – the numerical rank differential. If the higher-ranked rikishi wins its the positive of that differential; if the lower-ranked wins its the negative (e.g. M2 vs M9 is a score of +7 or -7 depending on which wins). Then just sum the whole lot up.
If the rankings say anything, the sum should be positive. The closer it is to zero (or even negative), the more competitive.
Interesting…I’ll bet theres a discussion to be had there with pundits of…..Another Sumo forum…Who’s prognostications are often..bleak, to say the least…
I just noticed (ok I am probably late to the game) that recent articles are filed under the category “Aki Basho 2022” :)
Ah, I will fix that.
Yay numbers! Here’s a random test idea that focuses on the whole division.
There are 42 rikishi in the top division, tallying all the ‘upsets’ by amount of ‘upset’ (e.g. if the bottom makuuchi rikishi, #42, beats the top makuuchi rikishi, #1, I give 42-1=41 points). Divide that by the number of bouts (including fusen, probably should exclude that, but meh, that’s extra time and this is just an experiment), to get the ‘upset index’. Kyushu 2024 had 713 points divided by 302 total wins in division, for an upset index of 2.36. Hatsu 2010 had 859 points over 303 wins, for an upset index of 2.83. What if any use/interest this may have I digress.
Yeah, pretty similar in principle to what I suggested. Should indeed give an indication of overall “strength” of upset.