Tournament Statistics (NERD ALERT)

We hosted a submission only tourney today. Probably the most fun thing we have ever done as a team. Several different area schools showed up and everyone loved it. We used the rules that were in place for the Gracie Nationals. The only difference is that we allowed all submissions at all belt levels. There were no kid divisions; only adults. 15 minute matches. If there's no submission, neither guy advances. No time limit in the finals.

We kept statistics. Here they are. Let's discuss trends, success/fail rates, etc...

What other stats could be kept and what could we do with this kind of information? It could re-shape the way we train! What do we train to finish with? What do we train to defend against and escape from? At what point in a match do certain submissions become more of a risk? What percentage of the time is the Rubber Guard successful to sweep with? Submit with? Imagine that there are 2:20 left in that match and you hear your coach yell, "Don't go for that Arm Triangle. It's only successful 14% of the time after the 10 minute mark, and the other 86% of the time it results in a guard pass or a reversal." (This only a hypothetical; in no way does it reflect compiled statistics.)

Oh, the possibilities!

From now on I am going to have someone recording in-depth statistics for each match we compete in at all events.

EDITED WITH UPDATED STATISTICS

http://s15.postimage.org/6gtlagf9n/April_14_2012.png
image upload

Holy Mother of statistical analysis! I'm loving the breakdown!

Hmmm… no oma-plata’s, slicers (other than ham sandwich), or spinal locks to name a few things that stick out. What is the stat separation on gi/no-gi here?

My love of jiujitsu & love of math are finally coming together...

lemme see dem vids bruh

This is way cool!! What stands out for me is the less than 1 minute super fast subs, Armbar/triangle. I'll be on the look out for that next weekend. Both defense and attacks. They can be done on a very fresh full power opponent. We will spend time this week drilling armbar/ triangle defense! Also,3 out of 4 RNC came in the second minute.

That is awesome information

Interesting that the finish due to fatigue came just a few seconds after the match had hit the average match time.

Quote:

---

Originally Posted by Brian Debes

Hmmm… no oma-plata’s, slicers (other than ham sandwich), or spinal locks to name a few things that stick out. What is the stat separation on gi/no-gi here?

---

great question. most of these matches are no gi. i can break it down further. give me a little time and i will get that answer for you.

as far as omoplatas, there were some used to sweep but no submissions.

Quote:

---

Originally Posted by Chris Garry

Interesting that the finish due to fatigue came just a few seconds after the match had hit the average match time.

---

Very interesting indeed. This was his second match. His first match was a loss by armbar (in the no gi bracket) in 3:31. He tapped due to exhaustion at 5:41. He had 8:12 of mat time before he was finished completely.

The Science of Jiu Jitsu!! Thank you for sharing this!

Quote:

---

Originally Posted by Tyler Woolsey

The Science of Jiu Jitsu!! Thank you for sharing this!

---

Glad you liked it, man. I'm looking forward to adding more data the next time, specifically success/fail rates on techniques.

One major problem I can foresee with doing something like that is codifying the language so that it's universal to all jiu jitsu practitioners/grapplers/fighters. For instance, how do you classify an armbar from the guard? Is it going to depend on the particular style of guard that was used to set it up? Do armbars set up from a traditional BJJ guard have a different success/fail rate than an armbar from Spider Guard?

However, with the approach that Eddie has taken of giving everything a name (even transitions) it's much less of a daunting task.

I'd be interested to see what path was used the most successfully at each level, and at which point a given path failed the most times.

Quote:

---

Originally Posted by Brandon Mccaghren

One major problem I can foresee with doing something like that is codifying the language so that it's universal to all jiu jitsu practitioners/grapplers/fighters.

---

Its all ones and zeros:)
Would be funny if your info was one day used for the 10th Planet entry in a robot jiu jitsu tournament

this is awesome.