Projecting a Champion: Part 1

One of the interesting by-products of my dive into Miyagino-beya’s silver medal squad was uncovering a potential new jewel in Enho. For a stable that produces very little in talent behind possibly one of the greatest rikishi ever to walk the earth (and let’s be honest, Miyagino-oyakata can be forgiven for that), it was curious to see a newcomer blast his way to a 7-0 yusho in the lower levels.

When looking to see whether this was at all something he shared in common with the current crop of sekitori as a possible signal of success, the answer was overwhelmingly yes. Of course, it is not the only signal, and as with many other sports, so much can go wrong (injuries, confidence, new trainer, etc.). But several top level rikishi put up multiple such records in their early (first year) basho and many more did it at least once. So this then begs a new question: out of the hundreds of men trying to break their way into professional sumo, is it possible to use this indicator to pluck names from down the banzuke – and if it is, when do we expect to see them bringing honour to their heya? If they don’t manage it, how much harder is it to then reach the top? This will be the first part in a series to attempt to try and figure all of this out.

First things first, let’s look at the most recent crop of 70 sekitori. We can use the March banzuke as a guide to figure out if there’s a story here because any makuuchi/juryo promotions/demotions aren’t going to massively change the calculus and the turnover from juryo to makushita doesn’t meaningfully affect our sample size (and we’re not only interested in who might make it to Juryo 14W someday).


What I’ve done here is split the sekitori into four categories: those who managed multiple 7-0 records in their first 6 full tournaments (post-maezumo), those who managed it one time, those who didn’t manage it, and those who didn’t manage it but were already in Juryo well before their 6th tournament (owing to entering the banzuke at makushita level as an amateur champion). This doesn’t tell the full story but it is somewhat telling that half of the professional ranks managed an early zensho. Of course, you’d expect the wrestlers who have been contending for titles their whole career to float towards the top, and so let’s see if that’s what we’ve got:



This confirms those suspicions. Makuuchi contains a higher number of rikishi to have put a zensho on the board their first year at least once (Kisenosato registered his first in his 7th basho otherwise this would have been more extreme), and as you’d expect it also contains a larger number of amateur champions who fast tracked their way to the top (and stayed there).

For the next parts in this series, we’ll start to look at how long it took these sekitori on average to reach juryo, and then start to look at the success rates of those who score this record and start to identify commonalities among them (kimarite, stature, etc), to model out who we might expect to charge up the banzuke soon and give us some more lower level candidates to track over the coming year. Again, this is just one of many signals – and there are many other intangible variables (stable, personality/confidence, etc) but it will be interesting to dig in and get an understanding of how impactful it is.

5 thoughts on “Projecting a Champion: Part 1

  1. Wouldn’t it make more sense to look at all the rikishi with early 7-0’s, not just the ones who ultimately reached the sekitori ranks? As it is, it’s a bit like concluding that the sun rising in the East causes it to rain, because every rainy day started with a sunrise in the East.

    Liked by 1 person

    • Yes and that’s the next step (thus why this will be a series). But before you can do that, it’s important to identify if there’s any signal at all to begin with. If we found that the top division were composed of a bunch of Hakuhos, you could likewise conclude that it’s not actually raw talent that’s important to the longevity of a rikishi’s career and their ability to reach the top level, but instead determination and ability to overcome losses and change up their game to meet that of their opponents.

      So the next thing we look at here is a larger sample of 7-0s over time, and if as is commonly held that only 2-3% of rikishi reach the top division, how do these guys then over index that, and what are their commonalities. I could have done the whole thing in one go, but it would have been a 4000-5000 word post, and there are going to be moments when there’s not much news going on so it makes sense to space them out.


      • My bad, I misinterpreted your final paragraph about the future instalments and thought you’d be sticking to the same frame of reference.

        IMHO, how meaningful 7-0’s are is highly dependent on the context. If one knows nothing at all about these rikishi except their honbasho records, there’s a pretty good signal there. On the other hand, at least for jonokuchi and jonidan nearly all those 7-0’s are foregone conclusions. Any halfway decent rookie is getting written up by the press before he ever steps on a professional dohyo, and the subsequent early fireworks merely confirm what everybody already knows at that point.

        In the Natsu jonokuchi division, the three most advanced guys (Enho, Tanabe, Fukuyama) combined for a 19-2 record, and both losses happened within that group (Tanabe to Enho, Fukuyama to Tanabe). That’s pretty much how it always goes, and jonidan this month won’t be much different for them in all likelihood. At those levels, the only other opponents who are a threat to them are rikishi who are in a comeback from a big injury and majorly underranked.


        • For sure – it reminds me a little bit of in baseball where you introduce a 22-23 year old guy to low or high A in the minor leagues and they steamroll through it but that’s not a massive signal in itself for whether they have what it takes for the top. It might be interesting also with these guys to start looking at their age vs. that of the division when making these accomplishments as well.

          There’s a real slippery slope here – especially in making that baseball comparison. I think sabermetrics have taken a lot of the romance out of that game. What I hope we can accomplish is introduce some analytics to forecasting while still remembering what makes the sport great is the individuality and unpredictability and the fact that it’s possible to make gains from one tournament to the next.


          • Re: the age factor…for what it’s worth, my personal impression from nearly a decade of prospect-watching is that promising guys are extremely likely to reach makushita by age 21 or earlier if they have started out of middle school (age 15) or high school (age 18). Consequently the collegiate rookies, being already 22 when they debut as pros, are pretty much playing catch-up from day one. That makes it hard to judge how much upside they have – by the time there’s sufficient data on them, they’re often either already up in juryo, or they have demonstrated that they won’t be getting there at all.

            In any case, my own prospect interest is more or less confined to the makushita division, because that’s where 90% of the interesting stuff happens. The college guys barely spend any time in sandanme and below, and the younger ones are so numerous and tend to spend so much time down there that it’s a fool’s errand to track them. I’ve got 60+ guys on my watch list elsewhere as it is. (About three quarters in makushita, the rest former makushita who have fallen back.)

            I’ve used the baseball comparison before, too – the whole tiered setup of both sports makes it very hard not to. 🙂 I used to worry about the “sabermetrics issue”, but not anymore. The key difference IMHO is that sabermetrics is ultimately about player evaluation for the purpose of team building (players are just building blocks in something bigger), while any such sumo stats work will be rikishi evaluation for its own sake. That limits the amount of demystification that can occur. And of course sumo doesn’t provide anywhere near the amount of competition data that baseball does, so any projecting is likely to be wildly inexact even at the best of times.

            Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s