This post is about a way to improve your development rate by devising game plans against fencers whom you have information about, but have never fenced against. It requires quite a bit of note-taking, but none of the steps involved are intrinsically hard. If your fencing is strictly at the recreational level and you are not willing to spend time outside of the fencing hall to get better, this is not for you.

As has been noted before, I have stated that it is easier to win over a specific opponent if has a plan for doing just that. Optimally, one has significant experience in winning over just that opponent, and he has not learnt anything from being beaten over and over again. In that case, one can simply apply what worked well before, no modifications needed.

However, reality is almost never optimal.

Instead, one has to deal with the non-optimal reality that one has in front of oneself. This reality differs from the optimal case in two ways:

  1. One is fencing against an opponent whom one has fenced against, and won over, multiple times before, but this opponent has in fact learnt something. (He might have been reading my previous blog posts!)
  2. One has not fenced against the present opponent before at all.

In this blog post, I will not further consider the first case. In the second case, the best way prepare a plan against the opponent is to consider other opponents who are in some way or another similar to one one is about to face, Those similarities could of course be physical, but what I am proposing below is another type of similarity altogether: most common mode of getting points.

There are three distinct ways a fencer can score a point in a fencing bout:

  1. Scoring a touch on a valid target area
  2. The opponent stepping behind his back line
  3. The opponent getting a red card

The last two cases are much more uncommon than the first one, and they do not inform specific tactical or strategic choices. Therefore, I ignore from now on in the classification system described below.

The points scored by actually touching valid target area can also be subdivided, and I do so by listing their level of priority. Note that this applies for all weapons, including epee. Priority does not apply when awarding points in a two-light situation in epee, but the different types of actions still exist in epee. The offensive actions, in decreasing level of priority, are:

  1. Point in Line
  2. Attack
  3. Taking of the blade
  4. Riposte
  5. Counterattack
  6. Reprise/Remise/Redoublement

Let us consider all those offensive actions with regard to epee:

Point in Line: This is not all that common, and when used, is used mostly to create distance and a momentary breather. It is quite uncommon that a fencer will score with PiL against an opponent who is at least minimally observant.

Attack: One of the major ways of scoring points in epee.

Taking of the blade: One cannot score this way without anything else happening, usually a riposte, or less likely, the opponent conveniently impaling himself on one´s blade.

Riposte: One of the major ways of scoring points in epee.

Counterattack: One of the major ways of scoring points in epee.

Reprise/Remise/Redoublement: One cannot score with these actions directly – one can only score with these actions if one first fails to do one of the above, is not scored upon by the opponent, and one then gets it right on the second or later try. So, points scored with these actions are indicative of both fencers trying to do something else, but being less than technically perfect in doing so.

So, we have three major ways of scoring points, and three that will be much less common.

Whenever one has three types of anything which together make up a given total, there is a quite simple way of describing that in a graphic manner – the ternary diagram. The idea of this type of diagram is that if A, B, and C constitute the all of the observations, then those three data can be plotted on a two-dimensional diagram, since there are only two degrees of freedom – A and B can change independently, but the proportion of the whole that is taken up by C is already decided. Example: if one knows that A constitute 10% of the observations and B account for 20%, then one already knows that C will account for the remaining 70%. Ternary diagrams are used for all sorts of statistical observations, and the concept is quite easy to grasp. If you have not come across it in your previous studies, I strongly recommend that you read the linked Wikipedia article.

So, the preceding boils down to that it is possible get points in at least seven distinct ways, but that only three of them (attack, riposte, counterattack) will account for the great majority of points scored. This leads us to a method of classifying fencers by the type of points scored in their matches. It goes like this:

  1. Observe matches and classify each point according to the types listed above. Do that for both fencers. This is best done on video, which gives the option to go back and study each fencing phrase again if necessary.
  2. For each fencer, note the number of points scored by the major types (attack, riposte, counterattack) and the other types. Do that for both points scored for and points scored against. If the total number of points for the other types is significant, then the conclusions drawn later on will be less likely to be of high predictive value.
  3. For a given fencer, add up the number of points scored by the major types from observations of several bouts. Do that for both points scored for and points scored against. Calculate the total number of points scored, both for and against.
  4. Take the data found in step #3, and plot where in the ternary diagram the fencer falls for the points scored by that fencer. You now have a ternary diagram with one point. Label that point with that fencer´s name. Repeat that in another ternary diagram, but with the data from points scored against. (Example: the fencer that you are studying has scored 15 attacks, 13 ripostes, and 8 counterattacks in the bouts that you have studied. His value for the attack corner of the ternary diagram is then 15(15+13+8)=42%. His value for the risposte corner of the ternary diagram is then 13(15+13+8)=36%. His value for the attack corner of the ternary diagram is then 8(15+13+8)=22%. The three values add up to 42+36+22=100%)
  5. Repeat steps #3-4 for all observed fencers, and plot them in the two ternary diagrams. Take care to plot each diagram in the right diagram!
  6. You now have two ternary diagrams, with a lot of labeled points. One of those diagrams – the one based on points scored for – describe the preferred types of offensive actions of various fencers. The other diagram – the one based on points scored against – describes the various weaknesses of those fencers.

Hopefully, you have done this groundwork for a whole lot of fencers, including those that you have not yet fenced against. If not you – your coach, parent, or club mate.

So, how to use the two ternary diagrams?

  1. Take note of where in the diagram the opponent for whom you are devising a plan is.
  2. Find other fencers close to that opponent in the diagram whom you have fenced against, and won over.
  3. Recollect which strategies and tactics that have worked against those fencers. Do that for all those fencers.
  4. Repeat steps #2-3 for the other ternary diagram.
  5. You now have a set of strategies and tactics that have worked against opponents that are similar to the one you are going to fence against. Choose the strategies and tactics that pop in many of those cases. Go through that list, and discard those strategies and tactics that directly can be discarded for other reasons.

So, there we are – a method on how to put together a game plan against a fencer that you have not yet fenced against. It does not require anything else than a significant number of observations, someone who is able to classify actions as either attack/riposte/counterattack, some basic math, the ability to punch in numbers into some statistics software, and a maintained list of what has worked for you previously. That is a fairly long list, but none of the items is all that difficult – they only require a bit of work. Better yet, a lot of the work can be farmed out to a helper.