Yay for Ura! First Win for the Pink Prince of Sumo
Let's do a profile of Ura!
Yeah, I’ve been falling behind in my sumo stats, but it’s tournament time! I’ve got to watch the matches every day (and I’ve been hacking up a lung as well).
On Day 6, Ura, my favorite, finally got his first win in the November 2023 basho.
Ura Rank and Weight January 2022 - Nov 2023
Using my record of ranks & weights going back to January 2022:
Ura has been hanging in the high-rank-and-filers for rank, but has trouble breakout out of that area.
As we’ve seen in his first few bouts of this tournament, the sanyaku just grinds through him, more or less. They’re used to his tricky ways, and he has a very difficult time beating them. He has an easier time beating lower-ranked wrestlers.
His weight has had some fluctuations lately, and I don’t know if the lower weight is easier on his knees. Many of the guys seem to get a strain if they get too bulky for their frames.
Ura’s Winning (and Losing) Ways
Let’s see his top kimarite for winning (and how often he loses to those).
It is interesting when you see such a variation from the top two kimarite — oshidashi is in line with the group, more or less, but his winning percentage with yorikiri is very low.
The difference between the two techniques is that with oshidashi, one is shoving the other guy out of the ring, and you’re going to lose contact with the other wrestler’s body as he leaves the ring.
With yorikiri, that requires having hands on the other rikishi’s belt, and having hands on him as you force him out of the ring. I think Ura’s size is such (being shorter and lighter, in general), that it’s difficult for him to use yorikiri.
I think I should do an analysis cross-referencing my height-weight graph with the winning techniques - I’m thinking we’ll see differences there.
Measuring how unique they are
Looking at the distributions of Ura’s and Tobizaru’s winning kimarite, I was wondering - hmm, maybe I should see how their winning distributions differ from the entirety of the sumo database.
So I decided to use the Hellinger distance for discrete distributions.
Okay, I know it looks complicated, but basically for each kimarite, I take the percentage it’s a win for the wrestler, take the square root of that, and then subtract the square root of the percentage that’s a win in the sumo database. I square that difference. Then I sum all those squares of differences.
The factor out front is a normalizing factor — the largest this number can be is one.
So I’ve calculated this for Ura & Tobizaru thus far:
Ura: 13.1%
Tobizaru: 13.7%
So far, Tobizaru is a little more unusual than Ura.
We’ll see how much others differ as time goes on.
