The title of this blog post is a pun – but you will only understand it if you are knowledgeable about the sport of baseball. If that is not the case, have a look at this.

Now that you have availed yourself of that knowledge, let me list the similarities between baseball and sabre fencing that make those two sports especially suited to statistical analysis:

  1. Only a few people are involved in each play/fencing phrase. There are only 2 fencers, and in the baseball case a great deal of the plays end with only pitcher and either batter or catcher interacting with the ball. Even in the more complex cases, it is usually so that at most 4 people get directly involved. That is in great contrast to, for example, soccer, in which all field players (20 people) are moving at the same time.
  2. Each play/fencing phrase typically is only a few seconds long. That limits the the number of different things that can happen to a manageable number. For example: in sabre, only 6 different calls are needed to describe the great majority of all fencing phrases (attack simultanee, attack touch, attack-no/contreattack touch, attack/parade-riposte touch, attack touch/contreattack, attack/parad/contreparad-riposte touch). Other, more complex, cases exist, but they are fairly rare and can for statistical purposes (at least in the first iteration) be lumped together in a miscellaneous category. That is in significant contrast to epee, where the length of a fencing phrase is only limited by the noncombativity rule.
  3. The various actions are distinct. A batter will either hit the pitched ball, or he will not. If two sabre fencers start being offensive from the get go, then it is either so that one of them gets the attack, other there is a simultanee call. Not so with many other sports, in which there is a continuum between two different actions, making classification difficult.

Of course, there is a big difference between the two sports: time. The down time between plays in baseball is much longer than the corresponding time between fencing phrases in sabre, making real-time notation in sabre that much harder. ut that is a fixable problem with video recordings.

So what does all this introduction lead to?

As far as I know, no one has attempted to study fencing matches on video and assign distinct, categorical, classifications to every fencing phrase in context where lots and lots of fencing matches are so analysed, and the classifications are defined beforehand. Of course, there are lots of coaches who have looked at matches on a case-by-case basis, in order to figure out some suggestions to their fencers for the upcoming match. However, looking at a few matches, selected on the basis that a specific fencer is present, is not good enough as a basis for a statistical treatment.

I therefore suggest that someone should do just such a information gathering campaign, based upon hundreds of sabre matches. Once it has been done, one can start doing a big correlation analysis, and something is bound to come out of that. If I am correct in assuming that it has not been done before, then whomever does for the first time will have rich pickings to find something new.

A very nonexhaustive list of what could be studied with such a correlation analysis:

  1. In cases where the lower-ranked fencer wins over the better-ranked one, what does the lower-ranked fencer do?
  2. What, if any, are the strategic characteristics of fencers who rise exceptionally fast through the rankings?
  3. Do sabre fencers fence the same way in 5-point poule matches as they do in 5-point team bout?
  4. In matches that start out evenly but then end unevenly (example: match goes to 10-10 with neither fencer having more than a 2-point lead before that, and then ends 15-10) is there some typical change of strategy that precedes the change?
  5. If a fencer is significantly better at the short phrases (phrases that start and end within the guard lines, with neither fencer retreating at any time during the phrase) than the long phrases (a significant part of the fencing phrase takes place behind a guard line, with both fencers retreating at some point during the phrase) – or the other way around – is there some characteristic of how the opponent fences that are good predictors of success?

So there you are, sport school students! A ready-made thesis topic! Define your categories, start filming, and get to work!