Making Statistics Work For You
For the most part, we learn the same moves in the same way. The instructor teaches a technique and then the students pair up and go through the technique. The big students and the small students all learn the same move, and often get to practice it on each other. For this reason we might imagine we’d see similar moves being displayed by large players and smaller players – and in many cases… we do! Armbars, triangles, rear naked chokes, cross chokes – these are all techniques you’ll see in any weight class.
The thing is, if you really get to looking at the number in Jiu Jitsu competition – I mean the real statistical trends – you start to see some huge differences. The insights I’m going to bring to the table today stem from a lot of number crunching while watching competition footage over the last 2 years, and some interesting conversations (and private lesson) with multi-time World Champion (and genuinely nice guy) Caio Terra, who competes at super feather and bantam weight.
The differences see in the trends of competition footage (or that many people can intuitively notice after watching a lot of high level Jiu Jitsu) between different weight classes in terms of technique and position frequency.
Initial Research: Black Belt Level Gi Matches and the ADCC World Championships
About 2 years ago I carefully watched 40 black belt level Gi matches. Some matches I had to find online (X-Combat has a lot of great videos available to anyone on the web), and some were played from DVDs of . Twenty of these matches were at the bantam and super feather weight class (the lightest end of the competitive weights), and twenty were at the heavy and super heavy weight class (the heaviest end). After breaking down the numbers, here’s what I noticed in this initial inquiry:
Here are the results:
20 HEAVY / SUPER HEAVY Weight Fights:
Initial standup at the start of match- 16 Guard pulls, 4 Takedowns
Takedowns (total)- 8
Guard pulls (total)- 25
Knee on Belly- 4
Sidemounts consolidated- 13
Mounts consolidated- 7
Backmounts consolidated (including 1 hook backmount)- 9
Back to the feet (scrambles/ref calls)- 17
1 Gi Choke from Knee on Belly
1 Gi Choke from Guard
2 Gi Choke from Mount
3 Triangle Choke from Guard
3 Gi Choke from Backmount
2 Armbar from Mount
20 ROOSTER (Galo) Weight Fights:
Initial standup at the start of match- 19 Guard Pull, 1 Takedowns
Takedowns (total)- 8
Guard pulls (total)- 36
Knee on Belly- 1
Sidemounts consolidated- 2
Mounts consolidated- 2
Backmounts consolidated (including 1 hook backmounts)- 10
Back to the feet (scrambles/ref calls)- 22
RNC Crank from Backmount
1 Inverted Armbar from Mount
4 Armbar Bar from Guard
1 Armbar from Knee on Belly
1 Armbar from Mount (before actually settling to mount)
1 Triangle from the guard (rolled to mount to finish)
1 Flying Triangle (when opponent has sat to guard)
The trends here are outstanding, and several real common threads seem to be coming through:
First, at the lightest weights, grapplers are overwhelmingly more likely to sweep than to pass in almost all contexts, and if a “pass” does occur it is likely to be a forced backmount rather than a sidemount or mount – which are almost never seen at this light division. In fact, heavyweight grapplers are nearly five times more likely to settle into a dominant position like mount or sidemount than are the lightest fighters – a trend that I suspect to be consistent well beyond the small scope of this research.
In addition, sweeps were almost four times as prevalent for the lightest grapplers when compared to the heavyweights. Again this is significant, as it seems as though the bottom man seems to have a significant advantage (in addition to taking into account our last trend) in terms of scoring potential, and top and bottom are much shorter lived at these lighter weights.
Recently I also aimed to crunch the numbers for the 60 kg division (around 145 lbs) at the 2009 Abu Dhabi Combat Club (ADCC) tournament. For those who don’t know, the bi-annual ADCC tournament is one of the most prestigious grappling events in the world, and it is the absolute pinnacle of no-gi grappling. This is a breakdown of 17 matches total. As a disclaimer it should be noted that not every minute of every match could be analyzed due to ADCC DVD footage – but I venture that its safe to say that the number presented here are indicative of the trends within this division:
-6 Closed guard
-10 Sweeps (not counting finals [too many])
-More than twice the number of guard pulls as takedowns
-There were more than twice the backmounts as there were all other dominant positions added together
-More backmounts that closed guards!
With these fighters being about 145 lbs, we’d expect to see slightly more side mounts and mounts, but still a heavy emphasis on sweeping and back takes. As the available ADCC footage seems to show, these trends did in fact hold for this 145 lb division.
It seems that this positional trend between the lighter and larger grapplers will hold for gi and no-gi, and with more research it will be interesting to see if
Theories Behind the Stats
So why is it that this particular trends seem to have arisen? Why can’t smaller grapplers seem to get to traditionally dominant positions like side mount and mount? We could pose that smaller grapplers (a) are stronger pound for pound and so are able to push opponents away with their limbs from these positions, (b) have more practice defending passes from huge opponents, and so are not as susceptible to a pass from a smaller opponent, etc… Developing these theories will also be useful in developing a more grounded view of where the BJJ world is and where it’s going. The most important thing to know now is that these trends are real, and that with more research, more certainty can be developed around them.
How Statistics Change the Game
Statistics provides an accurate and reliable source for interpreting data – and is useful in almost every conceivable domain.
Lets take finance, for example. If you are deciding on mutual funds and had to make your decision merely on what fund came to mind first or what fund looked best in the moment, your chances of selecting the right fund wouldn’t be as high as if you knew the performance of these relative funds over time.
In this was, statistics don’t dictate the decisions we make, but they help us make sharper and sharper decisions. In the BJJ world, this means determining the highest percentage submissions or techniques to focus on, find defenses to the most common attacks within your weight class, and / or finding potential loopholes in the trends that you might be able to take advantage of.
In the past decades we’ve seen an increasing use of statistics in sports like baseball and football – which have allowed teams to determine better strategies for specific situations and focus on the offensive and defensive strategies that work. As Jiu Jitsu’s popularity increases, we can only assume that this kind of data will become more and more available, and that statistics-tracking companies might come into the BJJ game as they have in such a huge way in sports like baseball, and as they are beginning to in the MMA world (www.fightmetric.com).
If nothing else, this aught to serve as food for thought for the serious competitors out there who are looking to smash their particular division. For me, this means watching all of the black belt level under 140 lb no-gi matches I can get my hands on and figuring out trends and how to use them to my advantage. For you, this might involve looking at the matches at your skill level and weight class, and determining what’s happening and what it means for you.
Check out this Black Belt Match