Sumo is a simple sport. Knock your opponent over or force them out of the ring to win. The only tool available to you is your opponent’s belt. Do you use it? This brings us to the traditional question of Yotsu zumo vs Oshi/Tsuki zumo. When push comes to shove, so to speak, which strategy do you choose? At the tachiai, would you try to lock in on your opponent’s belt and grapple, like Tochinoshin and Asanoyama, or blast away with powerful tsuppari, like Takakeisho and Abi?
Slapdown vs Thrust out
Takakeisho and Abi don’t really have the same strategy, do they, though? Takakeisho’s goal is to blow his opponent away. Before the suspension, Abi wasn’t doing that. He and guys like Aoiyama use those slaps to set their opponent up for the coup-de-grâce, the slapdown. As Abi’s comeback continues…
“Wait, Andy. Gotta stop you there. Abi is now at his highest rank, so I think it’s fair to say it’s not a come back anymore. Abi came back. Now, you may continue.”
Umm…thanks? So, as I was saying. Frankly, I was a bit taken aback by the forcefulness in Abi’s thrusts now. With Abi’s rush into sanyaku, we will get to see whether his strategy has evolved any, along with his strength. For now, I still have him clustered among the “Slapdown” wrestlers, and Takakeisho is in with the Oshi/Tsuki guys…
“Wait, Andy. Hold up. Clusters? What are you talking about?”
Oh, right. Sorry. So, I decided to cluster all of the wrestlers according to their apparent strategies. I call it an “apparent” strategy by the prominence of certain kimarite in their wins. I limited the data to bouts over the past two years. And I focused on the wins here, dropping data on their losses for a future analysis on weaknesses.
Since there are few great rivalries in sumo, featuring wrestlers who have battled over and over throughout the years, my attempts to improve predictions are leading me down this path: first classify wrestlers, and then see how they fair against wrestlers of that category rather than strict head-to-head comparisons. For example, if Takakeisho faces Abi, I want to see how Takakeisho does against this broad group of Slapdown wrestlers, and then I see how Abi does against strict oshi-tsuki specialists.
At first, I wanted to divide them into three groups: Oshi/Tsuki, Yotsu, and “Balanced” but while reviewing the Oshi/Tsuki wrestlers I decided there was enough of a difference between guys like Abi to warrant their own grouping, the Slapdown wrestlers. Unfortunately, there were trademark issues around the word SmackdownTM, for some reason, so I couldn’t use that. Lastly, I’ve also identified a small and controversial group of “trippers” among the yotsu-zumo practitioners. In sum, there are five groups and I’ll show the first group, the Slapdown Artists, in this article…the others will follow.
Who are the Slapdown Rikishi?
I’ve decided to filter these groups down to just the sekitori in order to keep it manageable, for now. Those sekitori in the “Slapdown” group are in the chart below. Three of the four main indicators that I’m using for these groupings. As I mentioned, I started with a goal of splitting wrestlers into yotsu-zumo and oshi/tsuki-zumo practitioners.
However, it became apparent that a large number of oshi-tsuki wrestlers featured hatakikomi and similar pulldown kimarite as a dominant tool. Rather than pure strength, these wrestlers main objective is to play the balance game. While several to win a plurality ore a majority of bouts using oshi-tsuki techniques, the sheer prominence of slapdown wins signaled the need for an independent group.
What’s interesting to me is how they’re not all in the Aoiyama mold of massive men. Rather than moving forward, we find these men frequently win while retreating or dancing on the tawara. You’ll also notice that Midorifuji and Takayasu may not belong in this group. They’re definitely on the edge when I plot their kimarite ratios. I might move Takayasu back into the “balanced” category but he’s on the edge there, too. Midorifuji is even more on the edge as he’s actually closer to being a yotsu-zumo grappler. But his katasukashi strategy is like a yotsu-zumo version of the retreating slapdown and I didn’t want to have a category of one.
Tonight, on NHK World, is a documentary about Hakuho:
20 Years of Glory: Yokozuna Hakuho
DESCRIPTION
In 2001, a 15-year-old Mongolian knocked on the door of traditional sumo wrestling in Japan. After 20 years, 69th Yokozuna Hakuho held many of the biggest records in professional sumo, including 45 championships and 1,187 career wins. How did Hakuho overcome tough competition and climb to the top rank, with so many unprecedented achievements? We look back on the highlights of the great grand champion and his deep affection for sumo.
Broadcast In: EnglishDuration: 0:49:00
In my time zone, which is Eastern Standard Time, it’s shown at 11:10PM today, and three times tomorrow:
6:10am, 02/27/2022 | NHK
11:10am, 02/27/2022 | NHK
6:10pm, 02/27/2022 | NHK
Andy, this is interesting, but if your goal is “first classify wrestlers, and then see how they fair against wrestlers of that category rather than strict head-to-head comparisons” then I don’t see how a category that groups Aoiyama, Midorifuji and Takayasu is going to be useful.
Those two are definitely the toughest. They’re real outlier for their groups, more so Midorifuji. I feel they move backwards a fair bit more than others….but “moving backwards” isn’t in the database. If the model ends up being crap, this criteria is the first thing I will re-evaluate.
@lksumo please feel free to correct me, but I see two main methodological errors here.
The first, which Andy alluded to in the post itself, was that some wrestlers bridge categories. I can understand the reluctance of having categories of 1, but perhaps that should have been the cue that hard categories weren’t as good as maybe even a graphic representation of a linear SVM, or at worst a Venn diagram.
The second, which I freely confess I have no solution to, is that there isn’t necessarily a bijective relationship between kimarite and wrestler “style”, given some kimarites are wins of opportunity rather than deliberate executions. That being said, such cases are probably infrequent enough that it doesn’t affect the general classification of rikishi, but often enough that hard categories would be a mistake.
The second issue is one I hope to account for with frequency. Endo seems to be, strategically, a yotsu-focused grappler. Sometimes, though, he pushes his opponent out for an oshidashi win. I would hope that over time and with more data points, his “style” would become apparent. Interestingly, he’s not as yotsu-focused as I had thought. He seems to be “balanced”, relying on a mix of yotsu and oshi techniques.
This is where something simple like sorting by slapdown frequency gets tricky. Tohakuryu and Aoiyama are definitely “hatakikomi artists” while others on the list are much less so. One could go in two opposite directions I think. Less quantitative: just use the “eye test” for grouping. More quantitative: try doing some sort of clustering on kimarite frequencies and see if tight groups emerge.
To me, Abi and Takakeisho are both thrusters first, though Abi alternates left and right thrusts to the face/throat and Takakeisho goes for 2-handed thrusts to the center of mass. Their plan Bs also differ, with Abi pulling backwards for a hatakikomi, and Takakeisho bringing foes down by dodging left and swiping with the left arm. Even if both were counted as tsuki wrestlers, I wouldn’t feel confident drawing conclusions about one from the results of the other.
I’d also want to separate both from the pushers who maintain contact and push opponents like a football sled (e.g. Takanosho, Hokutofuji, Onosho).