Training Feint Responses

[jBirch]

So I've been playing with the idea of teaching a student how to respond to actions that may or may not be feints. Current thinking is that a large part of the game of fencing is to analyse your opponent and try to predict which actions are feints and which ones are real. This is one of the major reasons why we teach multiple reactions to the same stimulus, so that the student can select a random appropriate response.

On the opposite side, we teach the student to make a feint and then immediately adapt to the response (or lack of response) from the opponent.

Obviously, the "acting" fencer has a huge advantage over the "reacting" fencer: whomever can effectively feint first, wins.

I recently came across an idea (no idea from where) that the proper response to a feint is a "half-response". That the fencer "mirrors" the action from the attacker: if the attacker only partially commits to their attack (ie// they feint) then the defender should only partially commit to their defense (ie// they move in a half-parry, for example).

I'm not sure about this idea, though it is intriguing. It would seem to me that the response would by necessity be late and therefore, inadequate.

Any thoughts? Do you teach your students to respond to feints or do you teach them to randomly select from a set of actions based upon the provocation?

James.

[keropie]

To some degree, this depends on the weapon being fenced, of course. But, things I've seen/we've done:

Foil: if A feints, B may make a 'false parry' leading A to deceive and finish. Now B parries, and ripostes.

A feints, B attacks.

A feints, B feints (this is my favorite tree): if A finishes, B parries and ripostes. If A pauses, B attacks, possibly avoiding a parry as A awakens. If A parries, B deceives and hits.

Of course, a lot of this depends on the feint, and the official. If A feints smoothly, sees B attack coming (into his preparation) and is able to continue his feint into the target, never really stopping his footwork, it may not have been a feint after all. So then B's (presumably trained) response of AoP is wrong, assuming the official sees the action as A's attack. So one of the things we train is adaptation. If the official says you're wrong, don't be wrong. This is one of the reasons I'm a big fan of feinting at a feint: it gives you a framework from which to assess the action. If A continues, you know it's his ROW, so do something about it. If A stops, hey, you might as well take ROW, right? etc.

In saber, it's generally harder for someone to get the call if they feint and then finish into an action begun against the feint. So then this might be a little less of an issue. But of course, a good fencer (in any weapon) can make a 'false' feint that is simply a continuous attack the whole way. Certainly many fencers can make a real/false feint that will conviently finish is appropriate, or avoid a parry if necessary.

My epee experience is much less, but obviously we're no longer concerned with ROW. When I fence epee (generally mediocrely), I am more looking for opportunities to hit, and trying to not get hit myself. Which means if I see a feint where the point is right below my guard, I must respect that, because it may well hit me if I attack, make a bad parry, do nothing, etc. However, if in the process of making that feint my opponent has exposed a good target, it may well be worth the risk if I think I've got the jump on him/her.

And some of that philosophy can work in all three weapons. I love it when people feint outside of my guard from lunge distance. This screams 'step in and hit me' to me, so I usually do. Sure, they have ROW, they even started first (giving them a tempo advantage in epee), but I feel pretty good about closing out their attack or at least forcing them into a difficult attempt to hit me and/or pulling their hand back to do so (losing them right of way or tempo, which ever they may have).

So, overall, yes, I teach my students to respond to feints, but not always the same. Ideally, their response will not be random, but based either on tendencies of their opponent's, their own personal predilictions, or subconcious data. If none of those work though, random is better than dear in head lights. And of course, ignoring a feint is a reaction, as long as you decided to ignore it. It's a great way to annoy people and make them make more commited feints.

[Allen Evans]

I recently wrote this:

http://home.earthlink.net/~allenevans59/LIES.HTML

About feints. I'm not very pleased with the page...it wanders all over, and the perspective (between fencer and coach) switchs in the middle. I think it should be split into two seperate pages, but I haven't had time to sit down and revamp it. I probably won't for another two or three months, so I decided just toleave it up until then.

Allen

[MUHSfencing]

very interesting idea - I recall reading about this somewhere (?) before...

this seems to fit in with the concept of parrying 'la finale' at the last possible moment. Any movement before then would be not a true parry but more like a preparation to draw a disengage if that makes any sense.

Also if I could tell that my opponent was feinting I wouldn't parry at all but rather counterattack. This 'false parry' idea could serve several purposes - determining the intent of ones opponent and setting up a counterattack...

[D'Art]

If possible, teach them not to react until it's too late for the opponent to uncommit* themselves. This may involve literally doing nothing, or it may involve doing one or several things from the same stimulus (effectively making them the one in charge). Sometimes it may just be that you get them to take a half-hearted parry slightly forwards while taking a half-step back. If they find the blade, they can still go on and make the riposte. If not the distance should be sufficient that they can still make a second parry or retreat without getting hit. How you actually teach this in a lesson environment without it being a fighting lesson where the coach actually tries to hit the student (assuming you don't mean top level fencers), I don't know, I haven't got to this stage yet, just going by my experience as a pupil.

HTH

[D'Art]

I'll try this again, seeing as I was stupid enough to press the wrong button after spend almost fully 15 minutes typing up a response earlier.....:S

Personally, I'd choose a variety of things. Sometimes don't respond until the feinter has gotten to the stage where they are over-committed, then react. Sometimes get them to react, but in ways the feinter wouldn't expect. For example, a feint to the weapon shoulder would usually provoke a circle-6. Instead, get the student to just lift their hand into a raised 6. After a bit of practice, this can be adapted into a beat attack on prep. Or get them to do step back and do nothing else. Sometimes, get them to throw in a bind or something similar. Just enough variation to give them options, but not enough to confuse them.

You could also get tehm to take a half-step back, while taking a parry slightly more forwards than usual. If the student finds the blade, fine - they are now the daddy in the exchange. If not, still fine - they still have enough time to either move out of distance or take a second parry.

These are all things that are best built up over the course of a few lessons, and walked through at a slow pace the first few times, and the speed gradually built up. Eventually, you should be able to actually try to hit them and not get close.

How long this will take, I have no idea. This is all coming from a student's point of view. None of my victims have had the pleasure of my lessons long enough to get to this stage. Yet.

[Durando]

James,

While I admire your rigor and creativity, I question the logic of what you propose to teach. Let's see if I can say why with some clarity.

My initial response is to teach students to feint correctly. That is to compose a touch in as realistic a fasion as possible. I think this really depends far more on tactics than upon technique. As far as epee is concerned, I've never been taught (or taught) false actions. Every action should be based on some reality. (Hence you have all these actions which commence with a potshot and then continue in a more sure fashion. A flick 2, followed with a counter six take, for example.)

I know you're not necessarily talking about epee, but the logic of determining "what could be fake" sounds like a very confusing approach to an individual lesson--it is almost certainly better leanred in a controlled bouting situation. What you want to set up is a kind of double blind trap where you tell one student, "Try a disengage in [whatever line]" and, after this has worked once or twice, you give a little advice to the adversary such as "set up the derobement (or even the counter parry)". Recognizing that a certain action is a false one is of no utility unless the response is immediate.

Teaching how to recon prep and making attacks into prep would be time better spent in my opinion.

[erooMynohtnA]

I think the perfect response to a feint is no response. If you know an action is a feint, wait to parry it elsewhere. Of course, that's not being practical, because there's no visible difference between a good feint and a real threat.

When I work with someone, I try to teach them to classify offensive actions into two sections: threats and non-threatening miscellaneous forward motion. If it's a threat, defend, if it's not a real threat (preparation), attack into it or ignore it. A lot of beginners have a hard time not trying to parry every action you make regardless of whether they need to, and I try to break that habit.

After a fencer figures out whether someone is legitimately threatening them or just waving his blade around in some sort of preparation, I add in another element: only parry something that can hit you. If I attack a student from out of distance, there's no reason to back up or parry, or if I attack a closed line, just wait for contact and riposte.

Based on that, a half-parry is a pretty good response, because it forces the attacker to either switch lines or commit to the line he's in, yet the defender isn't overly committed. I like to think of a feint as what your formerly threatening motion in one line was now that you're threatening another line. The defender forces a feint by parrying, because if they didn't parry, it would have been an honest attack.

I think you're right on about the main part of the feint/parry game being the ability to auger what your opponent is planning. It's certainly something I have a lot of trouble with, and I don't know how to teach that effectively. If you have any tips, I'd like to hear them.

In the vein of what you said about the defense being late, I wouldn't go so far as to mirror what the attacker is doing. I think if the attacker is doing a slow/fast tempo change with a slow feint and fast disengage, you'll only fake yourself out by attempting to mirror that tempo, because they'll have hit you by the time you speed up.

So, no, I don't teach my students to respond to feints. However, I do teach them to respond to threats in distance, which is what most feints start out as. If they know someone isn't trying to hit them, I want them to wait until that person decides to actually attack a line in earnest, then parry when the attacker is committed. One technique I use (and try to teach, but no one ever does it) is to just remain in my six on guard position and let the attacker "feint" all over the place until he parries himself on my six and I riposte.

[Allen Evans]

I was taught to make "half-parries" as part of an old style saber drill, in which a feint is met by slightly "covering" the line while retreating. This was 'monkey see/monkey do' exercise leading to a "false parry/real parry" drill: the same idea with a change of tempo (the false parry done slowly, the real parry done quickly).

You see this being done a lot in saber still, now that second is the new tierce in saber. Saber fencers will often retreat while making a slight head parry, forcing their opponents to try to score to the flank or under the hand, resulting in a devastating second riposte with truly frighting speed (you see Ed Korfanty use this on occasion).

I'm not sure that the answer isn't one of those horrible "it depends" situations. I might make a student do a false parry/real parry against a very fast, simple fencer. I might make my student ignore all feints except the final attack against a fencer with a long deliberate attack. Or, I might make them execute a false counter, and parry riposte, or feint in tempo with someone throwing out a lot of random feints.

In epee, of course, it can be totally random. Sometimes a feint should be met with a fleche. Sometimes, it should be met with a false parry/real parry and riposte.

I'm not sure that anyone has ever written about this rigorously. We all talk about feints and teaching our students how to feint, but hardly ever talk about what to do about them when they are done against you.

Allen

[jBirch]

Originally Posted by Allen Evans

I'm not sure that anyone has ever written about this rigorously. We all talk about feints and teaching our students how to feint, but hardly ever talk about what to do about them when their done against you.

Allen

Thanks Allen,

This much more eloquently expresses the problem I'm experiencing. How should I teach students to handle the feint from the defender's viewpoint?

Don't fall for it is, obviously, not a good answer, though the one most everyone likes to give.

I'm familiar with simply teaching the student to change up their response to the feint and to treat it like a real attack. I don't like it because it essentially leaves the defense up to luck and it seems like a hack.

The real key is to determine how one should handle a feint in a consistently safe manner. The half parry seemed ideal when it was first mentioned because it allowed the defender to maintain their freedom of movement in response to the action from the attacker.

It seems to me that there are two significant weaknesses to a feint:

1) It must be slower then a straight attack. It is therefor open to simple counter-attacks.

2) It must elicit an expected response. It is therefor manipulatable if you can "counter-deceive".

Any thoughts where to go with this?

Thanks for all the responses so far.

James.

[jBirch]

Originally Posted by Durando

James,

While I admire your rigor and creativity, I question the logic of what you propose to teach. Let's see if I can say why with some clarity.

My initial response is to teach students to feint correctly. That is to compose a touch in as realistic a fasion as possible. I think this really depends far more on tactics than upon technique. As far as epee is concerned, I've never been taught (or taught) false actions. Every action should be based on some reality. (Hence you have all these actions which commence with a potshot and then continue in a more sure fashion. A flick 2, followed with a counter six take, for example.)

I know you're not necessarily talking about epee, but the logic of determining "what could be fake" sounds like a very confusing approach to an individual lesson--it is almost certainly better leanred in a controlled bouting situation. What you want to set up is a kind of double blind trap where you tell one student, "Try a disengage in [whatever line]" and, after this has worked once or twice, you give a little advice to the adversary such as "set up the derobement (or even the counter parry)". Recognizing that a certain action is a false one is of no utility unless the response is immediate.

Teaching how to recon prep and making attacks into prep would be time better spent in my opinion.

Agreed. The question is really, "What do you do when your opponent makes a good feint which will either finish if you don't react or hit you in your opening if you do?"

James.

[Durando]

Originally Posted by Allen Evans

Saber fencers will often retreat while making a slight head parry, forcing their opponents to try to score to the flank or under the hand, resulting in a devastating second riposte with truly frighting speed (you see Ed Korfanty use this on occasion).

So the person actually making the feint is the person making the half parry. I think an emphasis on provoking a particular reaction, conditioning responses in an adversary is a good way to go. I really don't see how working on what is a feint and what is not a feint really works while on piste and, therefore, I can't imagine giving instructions to the effect within an individual lesson.

A good feint is indistinguishable from a real action. A poor feint is simply read as a "mistake" by one's adversary. Much easier to ask a student to capitalize on a mistake than to first classify what sort of mistake it is and then react.

Again, I am far too epee-centric when I say this, but feints in the unconventional weapon come in all shapes and sizes. Turning your hip slightly can be a feint. You can feint a fleche by brining your back foot forward in a certain way, etc. This sort of thing I think is best learned in footwork sessions. For example, ask your students to do a half retreat, advance-lunge and then ask them what that initial half retreat does.

[Durando]

Originally Posted by jBirch

Agreed. The question is really, "What do you do when your opponent makes a good feint which will either finish if you don't react or hit you in your opening if you do?"

Hi James,

I think my discomfort with this approach is that it cedes the initiative to the adversary. I understand you must be talking about conventional weapons as well. The way you pose the question is unanswerable. You're damned if you do, etc.

If this were in an individual lesson, I would ask, or provide, a sequence that changes the initial conditions of the touch. Of course you have to go back to tempo and find a solution that allows the adversary to think he's going to make the same touch again while you subtly change the equation. I think this has little to do with recognizing whether a particular action is a feint. Perhaps you would be better off trying to help your student recognize patterns. "He always begins with an advance, beat four..."

Sorry, I really don't mean to belabor the point.

[RITFencing]

Originally Posted by Durando

A good feint is indistinguishable from a real action. A poor feint is simply read as a "mistake" by one's adversary. Much easier to ask a student to capitalize on a mistake than to first classify what sort of mistake it is and then react.

I disagree on this point here. I think that an excellent way to provoke a reaction is to make a "mistake." So long as you can control the opponent's reaction and capitalize on it, then that "ugly action" can actually be a very effective fient or invitation.

EDIT: One thing that I should note here is that I do not like teaching this kind of action to new students; only to people who already have the control and sense of distance/tempo to recoginze the difference between the real action and the fake mistake.

[rcmatthews]

I would submit that following your opponent's feint, even halfway covering the line, is the wrong thing to do.

First, you have ceded to your opponent. You are now playing a game in which your opponent has complete information (more or less), and you do not.

Second, why should you react to something which you know is not real?

I can think of two alternative methods of dealing with this at the moment. First, you can use your footwork to draw the attack from your opponent. Using this method, it doesn't really matter how he feints, because you will be the one triggering his final attack, and hopefully where on the body that attack comes from. (Ways of drawing the attack would perhaps be better suited for another thread.) Second, you can work on having the student hold off on parrying until the last possible second before they take a parry. This helps in learning to distinguish what is real and what is not.

Originally Posted by Allen Evans

You see this being done a lot in saber still, now that second is the new tierce in saber. Saber fencers will often retreat while making a slight head parry, forcing their opponents to try to score to the flank or under the hand, resulting in a devastating second riposte with truly frighting speed (you see Ed Korfanty use this on occasion).

I am currently experimenting with using this same tactic, but with a step into my opponents attack. It generally draws the attack to 3 or 4, but I will admit that I have only tried it against basic level fencers. I think it has some merits (the forward moving version, I know the backwards version does).