High Percentage martial arts takes data from hundreds of fights, competition matches, and self defense situations to answer the question: What's truly high percentage?

 

I asked Facebook to Predict the Highest Percentage Submissions. They Were Dead Wrong.

I asked Facebook to Predict the Highest Percentage Submissions. They Were Dead Wrong.

Earlier this month, I published the first article in my white belt series. Among other things, I recorded all of the submission attempts and finishes in one hundred, IBJJF rules, white belt matches. Then I broke down what all the high percentage techniques were.

Some of these findings were surprising to me. Like the fact that not a single white belt pulled off a kimura from the bottom, something I had always thought of as a very common submission. But others confirmed what most competitors consider basic knowledge. Like competitors who scored first won 62% of their matches. It made me wonder just how many students would find the data unsurprising.

So on a whim, I popped into a Facebook group I belong to (BJJ Everything, with about 18,000 members) and asked them a simple question:

 For this article, I only tackled the first part of the question.

For this article, I only tackled the first part of the question.

 

Almost instantly, the responses poured in. Over one hundred opinions from white, blue, purple, brown, and black belts. Most comments were serious takes on the question, some went on tyraids, a few people argued with one another. In other words, a typical internet comment section.

And while there were some great, thoughtful responses, the conclusion was pretty clear: When you add up everyone's opinions equally, they are dead wrong. 

Of course, everyone's opinion isn't really equal. More experienced competitors are likely to give better advice. But in the world of social media, it can be much harder to judge this. Often the most vocal or the most articulate voices win out, not always the most qualified.  

This isn't to say that my data isn't infallible. It has a margin of error and limitations like all data. But even those things couldn't account for the disconnect. The Facebook community wasn't even close to the data. Here's what Facebook thought was the most successful submissions at the white belt level:

003.png

As you can see, Facebook had a wide range of opinions (these are only the top ten), but generally seemed to agree the the kimura, americana, and armbar from the bottom were the "most successful" submissions for white belts.

Now let's take a look at our numbers:

 These are subs shown as a percentage of all the subs that I saw in one hundred matches.

These are subs shown as a percentage of all the subs that I saw in one hundred matches.

A couple of things jump out here

  1. The actual top submission (by a huge margin),the armbar from the top, wasn't even in Facebook's top five.
  2. Facebook's top submission, the kiumra from guard, finished zero matches and was only attempted twice.
  3.  Facebook's second pick, the americana, had just three finishes out of eight attempts.
  4. Facebook's fourth and fifth picks, the armbar from guard and collar choke from guard, had the two lowest finishing percentages of any submission attempted at least ten times.

 

WHY IT MATTERS

Despite the fact that people have more access to knowledge about BJJ than ever before, there is still a tremendous amount of disinformation about what works for the highest number of people the highest percentage of time. We tend to confuse what might work for us with what works for everyone.

If you were a new student listening to the hive mind of your Facebook feed, you would be investing your time into what is, statistically, the least successful submissions. Maybe you could hone those techniques to become very skillful with them. But you'd be a statistical outlier. But why play the odds when you can put them on your side? If you're a white belt, build your house on a strong foundation: the techniques and strategies that are proven to work.

How the Math is Against White Belts and Submissions.

How the Math is Against White Belts and Submissions.

Deep Dive: Here's What the Data Says About Takedowns

Deep Dive: Here's What the Data Says About Takedowns