I'm not sure that the terms I'm using are standard, or that I'm explaining them well. Comments and feedback are welcome. Let me try to explain my terms, explain my assumption and then ask a question.

Attack Modality

My training has emphasized responses to attacks that intend to harm. When I'm the uke (recipient of technique), I try to attack with intent; when I'm the tori (recipient of attack), I ask my partner to attack with intent. There is nothing more frustrating than what I call "static grab" - where uke walks calmy up, and deliberately grabs my wrist and stands flat-footed and balanced, and waits. If Uke shows no intention to attack or to do harm, then many of my techniques fall flat. (I'm using wrist as an example - the same is true if my partner throws a flat-footed punch, or any other "mock attack"). It bugs me slightly, but we have to teach our ukes to attack authentically in order to defend authentically. Or so I have always thought.

Today, one of my female training partners commented that a "static grab" is a common attack modality against women.

Speculation on gender difference in attack modality.

What follows is explicitly speculation - I'm just trying to explore my assumptions and blind spots and to understand her comment (yes, I will discuss it with her further.) It may be that "static grab" is a way of expressing dominance over women (and, I suspect children). Talking this over with my significant other, we suspect that there may be two situations where men use static grab against women:

  • Completely static dominance display - female acquaintances have told me of men grabbing them by wrist or shoulder. The stories I've heard of this suggest that this isn't an aggressive attack; there is no intent to do harm. The man just wants to impose physical contact on a woman as a way of establishing dominance (this seems to be correlated with bar/party/ environments). I've got a couple of "friend of a friend" examples of this; one or more women of my acquaintance have related tales of men who, without invitation, discussion or transition, simply reached out and began fondling a breast. That may cross the line into aggression, but I'm going to draw an arbitrary line and include it here because there was no evidence of intent to cause pain or injury.

  • Imposing truth through dominance - Female acquaintances have also mentioned more aggressive grabs or restraints in the context of an argument "Listen to me", or "Don't turn your back on me". The intent still does not seem to be an explicit escalation to physical harm, but the level of effect is significantly higher, and I think both the woman and any observer would perceive this as a far more explicit threat.


This is an incomplete hypothesis - I haven't finished thinking about it. I'm excluding examples where the attack is obviously and explicitly intended to do harm, or to lead to a situation where the attacker can do harm. These cases fall into the "authentic attack" modality and the gender of the individuals is irrelevant

I may eventually include man-on-man aggressive dominance - my hypothesis is that when men express a dominance challenge against other men, they do it differently than they do towards women.

This hypothesis would explain some comments that I've heard from female training partners. But it is a hypothesis, and I'd be more comfortable if I could find some evidence that would confirm or refute the hypothesis.

If this hypothesis is valid, then I need to change my training slightly. I need to incorporate training against static grab and be aware of how the response escalates or de-escalates the conflict. I think that is much more difficult than the kind of training I do now which emphasizes attack with the intent to harm.

The question

Where could I find statistics on attack types by gender? Where could I find raw data that I could categorize by attack modality? Has anyone done this?


3 Answers 3


The FBI compiles some data but not as fine-grained as you want. Beyond that I think you're SOL other than looking at guesses.

My favorite such nonscientific approach is the "HAPV" (Habitual Acts of Physical Violence) idea formulated by Patrick McCarthy. He seems to describe things accurately in my judgment. That is to say, he alleges that the most common attacks in the USA are a haymaker to the head, a grab and punch to the head, and a few others (tackles, shoves, etc) and martial arts should tailor their training at least to some degree around those. (NB: different regions have different HAPV distributions.)

The other area of inquiry to look into is ape dominance patterns. Humans, like all apes, have rather ritualized forms of displaying aggression before engaging in socially-motivated intraspecies violence. These forms vary depending on culture and can come in several forms, but they have been studied.

However, with regards to static grab training, I think two things should be kept in mind:

  1. Static grabs aren't static. Don't train them as "he grabs your lapel and waits", but rather "he grabs your lapel to hold you in place for a big honking punch coming down the pipe" or "he grabs your lapel and violently shakes you"
  2. In my opinion, training focused around static grabs is less effective and interesting compared to addressing static grabs in the context of full-contact grappling (as in wrestling, judo, SAMBO, BJJ and so on). In the former, grabs are the starting point. This often leads to blowing the problem out of proportion, or making overly complex solutions to the problem. In the latter, grabs are dealt with simply and briefly as they come up in the broadly applicable scope of grappling.
  • Re: #1 - those aren't static grabs. Those are dynamic grabs; those are the core of our training curriculum.
    – MCW
    Commented Mar 11, 2014 at 11:43

I don't think a statistic will solve your problem. What would be the consequence for the training if you find such a statistic? Someone in your training group has experienced a static grab as an attack. Do you want to tell him/her "Sorry, your case is a statistical outlier..."

From what you've told dominance display by grappling seem to be a bigger issue for persons with a weak appearance than a "harming" attack. Take it seriously, your trainees do it, and will appreciate it if you will.

  • 2
    I agree with the idea that one should be generalized in their self-defense training, so that they can be able to handle a broad range of attacks. But I think it is also important to recognize that some attacks are much more common, and to make absolutely sure you're able to handle those kinds of attacks. I don't think anyone suggests only training to handle the top 3 or 4 fight modalities. Commented Mar 11, 2014 at 16:53
  • Agreed. I'd just say rely on personal experience, yours or your trainees, instead of statistics. That's all ...
    – user1778
    Commented Mar 12, 2014 at 6:27
  • 1
    Except that most people don't have a vast amount of experience with actual self-defense situations to be able to even know what is common and what is not. Police officers hear many first hand accounts of how people were attacked when they're taking down witness statements. That's why law enforcement personnel generally are much more reliable than your average Joe. And that's where most of these statistics come from. Aside from that security camera videos are making their way to the web, and you can peruse some of the fights they show on youtube. Commented Mar 13, 2014 at 4:14
  • Agreed again. I just don't think that's the point. Maybe your trainee or a friend of her has experienced somebody trying to keep her down by grappling. Would it matter to her if you can reference the statistic xy, which states good old fashioned beating ups make 99% of all attacks? Really try to get that situation from her point of view. And even if the statistic says something like this, is it valid? The number of unreported grabs should definitely be lower than unreported Glasgow kisses. If it is not in the scope of your training, don't train it. Don't excuse this decision with statistics
    – user1778
    Commented Mar 13, 2014 at 6:02

It's information which is sadly rather difficult to get hold of, but which is crucial for all martial artists and security professionals to understand. Given that some martial arts were designed originally to help deal with these most common types of violence.

This is an article which John Titchen recommends. The data was compiled and summarized/posted by a Shotokan practitioner who was looking for exactly the same kind of information.

John Titchen blog post on habitual acts of violence created 2005, updated in 2015.

  • This "Data has been gathered from a Home Office study group formed to investigate violence within modern society." However, there is no link to said study, its findings, or anything but vague references to it. Furthermore "Therefore when defending yourself, Show No Mercy." shows a candid attitude to the law in the UK. Just because you are attacked, does not give you carte blanche. Commented Mar 7, 2016 at 12:58
  • "most martial arts were designed originally to help deal with these most common types of violence" — even if this were true, which is highly dubitable, it is fairly reliable that habitual acts of violence change over time and space, therefore what an urban Japanese man in 1850 needed to deal with might be irrelevant to a rural Californian woman. Commented Mar 7, 2016 at 17:06
  • @Sardathrion - good point, I don't have a link to the original report. It's likely that neither are online at all. I've therefore updated the link to point at JWT's blog instead. Commented Mar 7, 2016 at 17:21
  • @DaveLiepmann it's fairly likely that an urban Chinese/Japanese man in 1850 would be attacked in a similar manner to an urban Chinese/Japanese man today, or indeed an urban English/American man today. Humans simply haven't changed so very much over time or space. Prevalence of firearms in specific countries is probably the largest change. Commented Mar 7, 2016 at 17:27
  • 1
    Human morphology doesn't vary much across the parameters I describe. Human culture, which strongly affects the particular manifestations of aggressive, territorial, and predatory behavior, certainly does. Commented Mar 7, 2016 at 19:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.