Where are the women fencers?

[eac]

In fairness, I have not ever had coaching from the likes of a Korfanty, a Burdan, a Luchetti or Ma or Gelman...or even a Smith or D'Asaro.

I think this is key. Having had coaching from coaches in that class and not in that class, the difference is enormous. Of my relatively small sample of the coaches not in that class, there was nothing any of them had to say on the strip that was of use to me.

But what made the causal link "clear"? I mean, how do you know you didn't just observe plain old correlation?

Well, I heard what the coach said, I saw the fencer do exactly what the coach said, and get one light.
Then, I asked the fencer later what he had been intending to do before the coach said to do that, and it was something else, which he had tried already several times with about a 50% success rate.

That's about as causal as it gets, I think.

[edew]

As a person who has been on both sides of the sport zone barrier, I think the most useful coaching help is correcting the perception of distance. As a fencer, it's easy to lose focus on distance when the legs get tired and people move slower. An "impartial" coach can see that you're too close or too immobile and will tell you to open up the distance or move a bit more. Hard to do that on one's own.

On strip coaching, most of my comments are done before the bout starts. "He likes to go fast." "He likes to parry." "This ref can't see anything happening in the box, so better to not get sucked into fighting there." After that, I'm basically cheerleading. True, that's mostly in saber.

I've seen non-saber coaches offer saber advice. Pathetic. It was causal and not correlative: the fencer did what the coach said, but did it terribly because it was way too complicated for saber. (It would be like asking someone in saber to make a feint-disengage cut, get parried, then make a compound counter-parry and then a delayed riposte. Not too unusual for foil, but way too risky in saber when a minor error would mean touch against.)

[ysbadadden]

As a person who has been on both sides of the sport zone barrier, I think the most useful coaching help is correcting the perception of distance.

I agree from experience on both ends of the relationship. In foil/epee, the most useful advice is "distance!" In saber, it may well be "smaller steps!"

[edew]

And of course, for all weapons, all ages and all skill level: "bend your knees!"

[ysbadadden]

and of course, for all weapons, all ages and all skill level: "bend your knees!"

qft...

[sabregrrrl]

And my favorite coaching advice: "Finish your attack!" Really? Is that what I'm supposed to do?

[Inquartata]

Well, I heard what the coach said, I saw the fencer do exactly what the coach said, and get one light.
Then, I asked the fencer later what he had been intending to do before the coach said to do that, and it was something else, which he had tried already several times with about a 50% success rate.

That's about as causal as it gets, I think.

Maybe possible in foil, or epee. Simply not enough time for planning of that sort in sabre, if the opponent so much as twitches in a way you didn't expect. And a coach cannot anticipate whether he will or not...

[eac]

Not enough *time* to plan like that in sabre? It seems like you have the same amount of between-halt-and-fence time. Maybe you mean that any given string of actions a coach could say wouldn't be sufficiently robust to variations in opponent actions? I can imagine that being true. I think because ROW is more elastic in foil, you can make things more robust by waiting for the conditions you want before moving to the next step in the sequence.

However, I still could imagine coaching being useful in sabre, if only in a more general way. For instance, in the sabre commentaries posted a while ago, they were talking a lot about how you had to be really in the other guy's face when you're on defense, or you get blown up. I could imagine a guy getting hit with an attack, and then the coach yelling "Get in his face more when you go backward!" and then that helping some.

What do you think about coaching in the one-minute break? Is that any better than yelling from the sideline?

[Inquartata]

All I can say is that the vast majority of outcomes from that that I have observed are either (a) coach advises, fencer continues doing what he was doing before, coach evinces anger and disgust, (b) coach advises, fencer does things which may or may not be attempts to follow advice, coach evinces either self-satisfaction or anger and disgust depending upon whether fencer wins or not, or (c) coach advises, fencer does something which does not resemble the advice given, coach takes credit for whatever it was anyway. Essentially it looks pretty random to me. Give enough advice and sooner or later there's going to be time when the touch and the advice seem to sort of coincide...but the causal connection is not proven by these coincidences, IMO.

I know that I have several times had the odd experience of having a coach congratulate me for taking his advice when in fact I had done no such thing...

[eac]

As an experiment, would you consider following Yuri (or a top sabre coach of your choosing) around in December in Div I in Pittsburgh, and listening to what he says and what his students do, and seeing if there's any correlation between them doing what he says and them winning?

[michaelheggen]

What do you think about coaching in the one-minute break? Is that any better than yelling from the sideline?

Yes, I think it is better, but maybe not better in the same way.

During the break:

• the fencer is focused on what the coach has to say;
• the coach can go into detail rather than just yelling things like, "More bounce! More bounce!" (heard at Nationals this year);
• the fencer can answer questions and offer the coach his or her own analysis;
• the coach can speak quietly instead of yelling for everyone to hear; and
• the coach can act to affect the fencer's mental and emotional state.

My personal opinion is that during those three minutes on the strip, my work as a coach and teacher is done. It is now the fencer's job to get out there and fence and to apply everything he or she has learned—and to use creativity to go beyond the limits of his or her training. It is that fencer's three minutes in which to succeed or fail—not mine.

But, that's just me—there is no shortage of fencing coaches in the U.S. who have a different philosophy—including many coaches of champion fencers—so I could be all wet.

-Mike

[Inquartata]

As an experiment, would you consider following Yuri (or a top sabre coach of your choosing) around in December in Div I in Pittsburgh, and listening to what he says and what his students do, and seeing if there's any correlation between them doing what he says and them winning?

Hey, I may be old, but that doesn't mean that time has lost its hold on me!

Seriously, even if I could do that, and come up with a valid methodology, and had close access---I mean, like standing next to him while he was coaching, which referees are not going to allow---how would I control for all of the exogenous variables?

And even then, you're talking about correlation. How would I get from that to causality? How much correlation would suffice to identify it as nonrandom?

[eac]

Actually, in my experience, in the later rounds when there are big crowds surrounding every bout, the coaches can't avoid being surrounded on all sides by lots of people, so you just have to get there early enough to be one of those people. Nothing illegal about it. And, if I were a coach, I wouldn't really worry about it, since if you're Yuri Gelman, it's not like someone is going to come along, listen to what you say on the strip, and then start their own club and take you down. You usually can't catch the stuff in the one-minute break, but whatever.

As for significant correlation, I'd say you want to estimate two probabilities. One is the probability of them winning when they don't do what he says; we'll call that P(W|disobedience), and then the other is the probability of them winning when they do do what he says; we'll call that P(W|obedience). This has to be compiled on a touch-by-touch basis, which I suspect even you are not either fast enough or geeky enough to do out on paper on the spot. However, you can estimate them in your head. If P(W|obedience) - P(W|disobedience) is greater than a specified threshold (I'd say 0.15, if you compiled it over three or more 15-touch bouts), then a fencer is justified in listening to Yuri. People might be too disobedient or too obedient to estimate it accurately, in which case the experiment would have to wait for another day to complete.

If you don't like my numbers, use different ones. I think the methodology is sound, though.

What exogenous variables are you worried about?

[Inquartata]

Actually, in my experience, in the later rounds when there are big crowds surrounding every bout, the coaches can't avoid being surrounded on all sides by lots of people, so you just have to get there early enough to be one of those people.

When they are standing at the end of the strip shouting instructions, perhaps. But not during the breaks, which IMO is about the only time advice could even have a chance to be valuable.

And what about the coaches who give instructions in languages other than English? No way to follow those, even if I could hear them...

What exogenous variables are you worried about?

Some of them I'm not going to mention, because they are useful when fencing and I am loathe to spread them around and make them less so.

But how can you account for things like simple inability to follow a recommendation? That is, the coach says do X, fencer tries to do X but fails to do so, or fails to score with it. How do you count that?

How do you account for physical variations, like height advantages or disdavantages, handedness, injuries, mood, etc?

[eac]

Obviously you have to throw out the foreign languages. It varies from coach to coach, so if there isn't content being communicated outside of the one-minute break, you also have to throw that out. In foil there are multiple top coaches from multiple top clubs who coach in English and have content outside the one-minute break, though. I bet you can find one who has a student in the top 10 in D1MS.

Some of them I'm not going to mention, because they are useful when fencing and I am loathe to spread them around and make them less so.

Do you really have trade secrets that significantly improve your results?

But how can you account for things like simple inability to follow a recommendation? That is, the coach says do X, fencer tries to do X but fails to do so, or fails to score with it. How do you count that?

You have to make a judgment whether the fencer is making an honest, clear, focused effort to execute the specified instruction. That's the 'obedience/disobedience' independent variable. The case you said would either be lose/disobedience or lose/obedience depending on your estimation of how much the fencer was trying to do X.

How do you account for physical variations, like height advantages or disdavantages, handedness, injuries, mood, etc?

Height advantages, handedness, and injuries are the coach's job to account for. If the coach says to do something, it damn well better be something that is going to work within the constraints of the situation. If the fencer tries to do what the coach said, but what the coach said would only work if the fencer were a foot taller, that's on the coach. That's another lose + obedience case.

In my opinion, it's on the fencer to control his mood. If you're in a DE and your coach is telling you to do something that would work, it's up to you to control your feelings and focus. So if the coach says to do something, and the fencer can't do it because he's in a bad mood, that's a case of lose + disobedience.