The Myth of the Untrained Stranger

I don’t think I’ve ever been in a real conversation about the use of martial arts for self-defense without someone adding a stipulation about “but against an untrained attacker…”. One of the assumptions I’ve encountered a lot relative to the use of martial arts for self defense is that of the surprise attack by some random guy with no training or experience. I don’t really think that’s a valid assumption, let alone a reasonable one.

TL;DR: You’re unlikely to be attacked in general. If you’re unfortunate enough to find yourself in such a situation, it’s likely you’ll know your attacker. Whether or not you know your attacker, you should try to be honest about the people you associate with and the community you’re in if you’re making assumptions about a “normal” person’s reaction. For me, that means assuming everyone knows at least the basics of grappling. Continue reading


Aliveness, with respect to training, is the notion that the degree of efficacy of training is related to the degree of realism in the setting and circumstances of the training. That is, training against a resisting opponent makes you more able to apply a technique while an opponent is resisting in a live context (not training).

Aliveness is a spectrum. On one end you have complete compliance. Tori charges up some ki and blasts a compliant uke who falls of their own accord. On the other end you have actual experience. Nothing will prepare you for an MMA fight as well as being in an MMA fight. Somewhere between those is everything we do in training. It’s important to note that the far side of the spectrum, the most realistic side, depends heavily on what the goal is. Being in an MMA fight won’t prepare you to stop a mugging necessarily. Some skills will likely transfer over, but stopping muggings will better prepare you to stop a mugging. Whatever your defined end-goal for training is becomes that far side thing.

Just like engineering doesn’t really exist without physics which doesn’t really exist without math, I don’t think higher levels of aliveness training would exist without the lower levels having done some research first. At some point someone in Japan was dissecting corpses to see how tendons and ligaments connected tissues and bone. Many of the old jujutsu masters were the equivalent of orthopedic surgeons. This dissection and experimentation on the ultimate complacent opponent let them test and discover the directions and angles which give efficient leverage against joints. That then gets tested on complacent (but alive) individuals, then with slightly increasing resistance up to the point of what we consider “full resistance”. We’ve all done that – you play around with something, when it seems like it will work you drill it, then you try it in rolling to see how it fairs there.

I don’t think I’ve said anything controversial up to this point, but I think we now at least have a common language for the real topic at hand: what degree of aliveness are you really doing, and what degree do you think is the minimum acceptable for training in preparation for your goal?

If we were to take two schools training judo – one for sport, one for recreation – the sport school would likely have students who are better able to execute throws in situations against resisting opponents. They’re more likely to have intense randori sessions. They’re more likely to have competition experience where they’ve had to figure out how to throw someone who’s resisting. One school is clearly more likely to produce an Olympian than the other. For the goal of sport judo, you want something that mimics those sport situations as close as possible. However, if the goal is self-defense, it could be argued that the recreational school is a better choice. Even though it has a lower degree of resistance in training, not having to mimic the conditions of sport rules means not having to worry about gripping restrictions or limitations on grabbing the legs. Honestly, it’s a balancing act of assumptions. If I can throw a competition judoka, I can probably throw anyone, but if I train without restrictions I’m probably better equipped to deal with my partner not having to give me a certain grip.

Many people will criticize the martial arts which have lower degrees of realistic training. For some minimum degree of aliveness, I think it’s reasonable to call bullshit when a martial art advertises itself as being effective in an untested situation. Like, ki blasts are pretty much never going to work on someone from outside your class who hasn’t been conditioned to fold when you yell at them. Aikido has a pretty bad reputation in this regard. It’s largely seen as working solely on compliant partners, being often criticized as a form of self-defense. When an aikidoka is paired against a wrestler it seems to always looks incredibly one-sided. I’d argue that the two should have vastly different goals, and that those goals should be taken into account when determining the efficacy of techniques and level of aliveness appropriate to training.

Let’s pretend I am the head bouncer at a bar, looking to hire a new bouncer and two individuals have come to me – one a nikkyu in Aikido, the other a blue belt in BJJ. Let’s assume they’re both about equally fit – similar heights, similar weights. One candidate likely has experience with wrist-locks in a martial art that professes non-aggression. The other likely knows better how to fight for MMA or theoretical self-defense scenarios. I’d hire the Aikidoka. I’ve known a lot of bouncers who bounced for many years. Few of them have had to be in a fight. Almost all of them have had to escort a drunk out of the bar without hurting them. Techniques like the Devil’s Handshake, Bouncer’s come-along, a basic hammerlock, or a similar basic wrist lock tend to be the favored means to do this. The Aikidoka is likely to have a better base to be shown these. This does assume I’m not running some Roadhouse-style dive bar where fights break out with a high frequency and kicking your rowdier patrons in the face doesn’t seem to carry a risk of lawsuit, to be fair. It’s betting escorting someone out is more likely than having to fight someone. But under those circumstances, I’d say Aikido is the more applicable of the two. Now, if I were in the business of managing MMA fighters and I got the same two applicants, I’d pick the BJJ blue belt. In that context, I want the person who’s more likely to win a fight, and for that context my bet would be on BJJ.

Everyone has a set of assumptions about their end goal(s). If you assume you’ll be in a situation where your partner is more or less compliant (too drunk to fight back, hasn’t noticed you because they’re engaging someone else), then there’s no real problem with polishing your compliant partner techniques. Problems really arise when we project our assumptions and end goals onto other arts, or when other arts are not internally consistent between their assumptions and training methodology. If someone training BJJ assumed that BJJ would prepare them for gun violence, but never trained situations with guns, there would be just as much of a problem as someone doing Aikido assuming they can step to the side of a shot from a trained wrestler. If we assume a majority of fights go to the ground, of course a martial art with no groundwork looks bad – but it’s likely not an assumption they hold. Almost none of us are training in perfect simulations unless our goal is easily reproducible like sport. Almost all of us rely on some level of assumption. Many arts have different assumptions from each other, and that’s what you really need to get at to evaluate someone’s training – what are their assumptions and why or why not might they be valid?

Variations Matter

Early in my judo career I couldn’t do uchi mata to save my life. I have relatively long legs, I’m just about 182 cm (pretty much 6’0″), so it should have been a bread-and-butter throw. Tanner had no problems with it. He would just grab and before I knew it I was on the ground.

The variation I first learned was what I consider “traditional”. You throw facing forward, your leg lifts up making contact along theirs more or less the whole length of the leg, and it’s the hand pull and twisting of your hips that finishes it. Later I’d learn that the reason it wasn’t working is that I didn’t have uke’s arm pulled out in front of me enough. I also learned how hard it was to get that much pull for me because even to the limit of what I’m physically able to do, it’s not as much pull as the coral belt who was teaching at the time wanted. If I’m in just the right configuration and the person has committed themselves forward already, I might hit it. Maybe.

The second variation I learned didn’t work any better. It’s the “hip version”. It looks about the same, but you actually attack the opposite leg – making thigh to thigh contact on the far leg for the lift – and then cut across to the nearside leg during the finish to cause the rotation. I can kind of fake this one by starting with a hane goshi and then cutting the leg across once they’re lifted, but I still am not quite sure what detail I’m missing that would make this one work the way it’s supposed to. Play with it as I may, only in a kata setting have I ever been able to hit it.

The third variation, which I hadn’t learned until only a couple years ago, is a lower uchi mata. You make ankle to ankle contact rather than whole leg contact, and are facing more toward uke rather than the same direction. I’d describe the action as an ankle pick with your leg – pulling to the rear outside corner instead of the forward outside corner to finish. This quickly became a staple throw of mine. It just works for me. Even if uke has shorter legs so I can’t get low enough to get my leg in, being able to pop theirs up usually lets me do some throw. Having said that, I’ve tried to show it to others, and they still prefer one of the other variants.

Personal preferences and slight variations in the application of a technique are important. The core set of techniques work, but sometimes finding the right variant of a given technique is what’s really necessary to turn it from hot garbage into a staple. Everyone has some variant on the notion that you can ask four black belts how to do one technique and get six different answers. If you feel like you just absolutely suck at a technique, but you want to do it better, it may be worth a try to see if there’s a different variation you can use instead.


Juji Gatame Variations

Ude hishigi juji gatame – the cross-body armlock – has two major configurations in judo. In the first, one leg is over the head and the other behind the back. This is the one that Catch refers to as the “Japanese armbar”. It’s the first one I learned as being the “traditional” way to do it in judo. The second configuration is one leg over the head and the other over the body. I’m pretty sure this is the default in BJJ.

Neil Ohlenkamp demonstrates these two variations in his JudoInfo article on the technique.

The first configuration is generally more applicable in a context where you’ve just thrown uke, especially with an arm or hip throw that places them in front of you in an almost knee-on-belly position already. It’s also good from knee-on-belly for the same reason. It’s faster from here because you don’t need to get the leg over the body. However, if you’re doing this variation you want to be off to a 45 degree angle from their body instead of perpendicular to them. You should actually use the foot behind their back to push slightly and the one on their head should be biting in to make that angle as it lifts their shoulder up, reducing the amount of give and preventing them from being able to sit up into you as an escape. Pointing the thumb toward their hips (as opposed to straight up or towards their head) helps prevent a Hitchhiker escape and will tighten the lock. Some people like to hook into the armpit with the foot over the head. The big reason is that it prevents the escape where they just push the leg over their head. I’m not personally a huge fan of that and as Mike had noted the other night at class being on the head helps keep more pressure in the position. It’s nice for a transition to reverse sankaku though.

The second configuration is better from mount or closed guard because you don’t need to clear the leg to the other side of the body, and in my experience, is probably the better control position. The Spider Web from here is often easier for me, but it’s also a nice position to set up a bicep crush or jigoku jime (which has apparently been renamed to “Hebrew Necktie” by 10th Planet when done for no gi? This is an old technique, the no gi variation was up on LapelChoke back when that was around). I’ve always been taught for this configuration to point the thumb straight up to the ceiling. Switching to sankaku here, I’ll thread the leg that’s over the body through the arms and over the shoulder. Almost everyone sits up thinking that I’ve given them the escape, but really I just had a hard time breaking their grips and decided a triangle would be easier. You can switch the legs (threading the leg near their head through their arms) for the same reverse sankaku you got from the first configuration, but again, it’s more work than threading the body leg.

So, the big differences? When you use it (KoB vs mount), what angle you want (45 vs 90), and the hand rotation (toward belt vs straight up). Everything else is situational so it’s not as clear cut. You can always throw your leg over the body one way or the other to swap between them as you need too.

On Kesa

How I learned kesa gatame in judo is almost identical to how Steven Koepfer shows it. Both legs straight – one at nine o’clock and one at twelve o’clock so that you have dispersed weight and a bolster to stop bridging in either direction or being able to get leverage to take you backwards. Mike was showing for no gi a variation that looks more similar to how Joel Bane demonstrates – one jack leg, one drive leg. However, we were still doing head-and-arm, not kuzure kesa.

Mike’s variation is a tremendous amount of pressure. If you’re not used to being pinned by bigger guys, if you don’t know how to control your breathing a bit, before he even pulls up on the head for the added leverage into the chest you may feel the need to tap just from chest pressure. Using one leg to drive in and one leg to lift your hips off of the ground makes that pressure. However, doing so means that both legs are pushing your body backwards. Without the gi, that doesn’t really mean anything. It just means you’re getting free pressure. They’d have to get the far (non-trapped) arm in front of your face/body to push you backwards in no gi. With the gi, it means they can grab the lapel and do the sit-up escape.

Building pressure with the variation I’m used to necessitates a strong pull on the arm and sinking your hips forward. Your hips should still be off the ground, but subtly low; the lower they are, the most weight uke is holding. Winding in and driving your hips forward still won’t generate as much pressure as the jack/drive system from the other though.

Typically, what we do in no gi is equally valid in gi, but not vice-versa. An armbar, a choke, even a shot you would use in no gi can be used in gi with minimal modification. But a lapel choke, a sleeve throw, or a grip relying on the jacket won’t work in no gi. In this case, the gi version works in both contexts; it’s just potentially suboptimal in no gi given that there’s a better way to generate force there.

Judo Mat Culture

My training partners aren’t paid to show up and train with me. When they lend me their anatomy to practice grappling on, they do so with the trust that I’m going to be careful. I trust them to do the same. Each person has the right to refuse any training partner. Someone who injures others due to recklessness, ineptitude, or intent runs the risk of having everyone refuse to be their training partner. Even if it’s accidental, if you break all of your toys you won’t have anything left to play with.

Judo, at least as I have experienced it, has that mentality of taking care of your uke. If I’m applying a technique and my partner gets hurt, it’s my fault. If they didn’t tap before I applied enough pressure to cause an injury it could be any number of reasons – they didn’t recognize they were in a submission, I didn’t give them a chance to tap, my control/angle was bad so I had to crank it when they felt safe – but I should have had enough control to release the submission before damage was done. It doesn’t matter if it’s drilling or randori. I used to be told something along the lines of “any asshole can hurt someone by using a ton of force. Grappling lets you control not only when, but how and to what extent, if at all, the person you’re working with will be hurt.”

Every time that I have caused an injury to a training partner it has been unintentional and I have felt responsibility to help, as well as to be more careful in that situation. A lot of stuff happens when someone is injured. They’re probably pretty angry, understandably so, at you, at themselves, at the pain in general. They might not want you around, but it’s my opinion that you should still make sure they’re okay, offer any help/supplies you can – including escorting them to urgent care or the emergency room as necessary. The group that I do judo with is phenomenal about this. It’s hard to stay angry at someone or question if it was an accident when they show earnest concern. When they step off the mat and don’t go back to practice until they’re positive that you’re going to be alright. I’ve seen a fair number of injuries over the last nine years. I have seen individuals who others will refuse to work with. I have not once seen a situation where someone took responsibility and tried to help and was spurned for doing so after the dust settled.

I haven’t really seen these practices in BJJ, at least not widespread. For some reason “he should have tapped” is an acceptable answer to someone getting injured, even in recreational clubs. As noted above, you won’t always have the chance to tap – your partner might crank, they might recklessly apply force in a position where you’re not yet able to tap, you might not recognize the danger. To be blunt, I don’t think there’s any place for “he should have tapped” in a recreational club. If you’re in a competition environment, somewhere that you’re training for fights and prize money, train as you see fit; but the club that’s simply training for fun, I can’t imagine such circumstances are acceptable.

That’s not to paint everyone in BJJ with the same brush, there are a few really awesome individuals who I’ve met and have the mentality of safety and wanting to take care of their partners. The guys who when you get injured, even if they weren’t the one working with you they’ll come off the mat to help you and at least make sure you’re okay. Those are the guys I want to work with. That’s the kind of mat culture I’d want to see cultivated.