Fencing where results, but not fencers, are handicapped

[PeterGustafsson]

Hi!


No doubt you have been in situations when there is a lack of possible training partners, and those available are of considerably different strengths.

What to do in order to make it competitive and interesting?

One way is to limit what the better fencer may do, as in restricting his valid hitting area, or disallowing specific actions on his part. However, there are drawbacks with approach, and therefore I intend this thread to specifically only to deal with another approach - handicapped (as in golf) results.

In this approach, the weaker fencer starts with a points standing better than 0, but both fencers are free to do whatever it legal in ordinary fencing.

We have used this approach in my club (due to lack of evenly matched fencers) and the question arose:

If one has found reasonable handicaps in the bouts between A vs. B and A vs. C, what handicap should one start out with in the bout B vs C?

The simpler, but related, question is:

If one first tries the handicap n1(AB) in the fist bout between A vs B, and it is shown to give a lopsided match, which should one then use in the next bout - n2(AB)?

The simplest approach to the first question is to simply add together the handicaps (with correct signs, of course!). However, that is not good enough. Assume that fencer A is considerably better than B and the best handicap between them is found to be for B to start with 8-0 in a 15 point bout. In other words, n(AB)=8. Furthermore, assume that fencer B is equally dominant over fencer C, n(BC)=8. Adding would give C a starting standing of 16-0 against A, which is obviously useless.

My solution was to instead consider the handicap as a measure of the fencers relative point-production speed. In the 8-point handicap scenario, this would mean that the better fencer is 15/7 (=2.14) times as good. So, if fencer A is 2.14 times as good as fencer B which in turn is 2.14 times as good as fencer C, it is good initial assumption that fencer A is 2.14*2.14 = 4.59 times as good as fencer C. So, if fencer A is to make 15 points, the handicap is to be set such that fencer C only needs to make 15/4.59 = 3.27 points in the same time. Therefore, fencer C should in this example start with a handicap of 12-0, or in other words n(AC)=12.

The approach can also be used to recalibrate handicaps between two fencers. Assume that fencer A is better than B, so the start out with a 9-0 lead for B. However, the first bout ends with a B win of 15-10, so the handicap was a little to large. In this example, fencer B scored 6 while fencer A scored 10 points, so A can be assumed to be 10/6 (1.67) times as good. If they are to end up equal in the scores, the handicap is to be set so that fencer A scores 15 points, but fencer B only needs to score 15/1.67 (=9) points in the same time. Thus, the next starting score between them should be that B starts with a 6-0 lead. (This is the n1(AB) -> n2(AB) question.) I have found this to give a fast convergence.

I already hear the retort: But the fact that A is usually better than B and B is usually better than C does not always mean that A beats C! Yes, I am perfectly aware that winning potential is not an obligately transitive function. However, this is meant to design a system which on average gives as good results (for AC-handicap) as is possible, given only the input of the the AB and BC handicaps, works equally well for any set of A,B, and C, and does not involve any guesswork or fudge factors.

Any better suggestions? Other thoughts? How do you deal with fencing nights in which there are no two fencers of roughly equal capability?


Have a nice time!

Peter Gustafsson

[Hauptman]

Most sports that use handicaps are individual sports, such as golf or bowling. Handicapping a competitive sport is extremely difficult because the outcome is heavily influenced by the interaction of the two athletes and doesn't just rely on their individual performance.

Honestly, I don't think you'll come up with a system that will produce valid results except over a large number of bouts to average out the variability.

To test this hypothesis you could get a number of fencers to compete in a fencing "league" over the course of a season and see what the results are. I'm sure you'll be able to devise a system for the second season that should be reliable over the course of a season, but probably not very reliable over a night or even a month.

[Mr Epee]

Speaking from my own experience, I have never found the handicapping of bouts via points to be a particularly beneficial practice exercise.

In reality, it only simulates the experience of accidentally allowed a less skilled fencer to have jumped out to an early lead.

Certainly, a handicap system may occasionally add to the element of "unpredictable outcome" which his highly desirable for all sporting activities, however, the nature of fencing is such that not all actions executed by a fencer are created equal. The result is often the repeated execution of "scrub killer" moves instead of the more valuable experience of progressing through various tactical solutions implemented with subtle changes in technique.

Handicap fencing, in my experience, leads to the more able fencer sticking his/her nose in the water hole and not coming up until the maximum score in achieved. Simple fencers are simple puzzles.

Yes, the more able fencer can take a disciplined approach to employing a variety of tactics/skills.... but if a disciplined approach is required regardless of the handicapping of the bout, then the fact that this bout is handicapped becomes superfluous.

Also, the situation often arises where the less able fencer knowing that only a handful of touches are required, resorts to unsound actions trying to score the "lucky" touch.

- MrE... who is clearly board with various political discussions.

[tchwojko]

Originally Posted by Mr Epee

- MrE... who is clearly board with various political discussions.

Pining for better discourse?

[tchwojko]

Originally Posted by PeterGustafsson

Any better suggestions? Other thoughts? How do you deal with fencing nights in which there are no two fencers of roughly equal capability?

What about team fencing? You could do teams of two instead of three if there are insufficient numbers.

Fence a different weapon for a change of pace?

Present a variety of handicap choices and let the pairs figure it out?

[Mr Epee]

Originally Posted by tchwojko

Pining for better discourse?

At this point, I'm almost willing to discuss the possible value of a Tactical Wheel Bush.

[edew]

I do certain types of handicapping to encourage (or discourage) certain actions. At the end of the day, no one really cares who won or lost a particular bout.

Some examples would be, hits made with a beat-attack are considered off-target (for those who habitually beat-attack). Or, any head cut (in saber) is considered a miss. One-lighter head cuts earn no points, two-lighters automatically give the point to the opponent.

For foil, I try to encourage proper set-up, so setting up a counter-parry, counter-riposte earns points for the better fencer (the fencer needs to learn how to encourage the opponent to make a parry AND A RIPOSTE, which is usually too difficult for beginning level fencers: they often parry and not riposte).

Another handicap is to offer two points if a fencer can successfully overcome a particular action. Suppose fencer M23J attacks and Bz94 makes a septieme-type parry (it's not proper because, hey, they're just starting) with a riposte. If M23J can, when it happens next disengages around the septieme parry, he gets one point. If he actually scores, he gets another point.

Recognizing situations (i.e., problems) and devising a movement to get around it (i.e., solutions) is a valuable component to the sport, right? So I try to encourage that, even if it's not fully successful. Point control is an easy task to teach. Just drill, drill, drill. Manipulation, setting up traps, seeing problems and opportunities are harder to grasp for many people, especially kids. So I try to encourage that by giving points just for doing the right stuff, even if it's not executed properly.

[tchwojko]

Originally Posted by Mr Epee

At this point, I'm almost willing to discuss the possible value of a Tactical Wheel Bush.

You went against the grain of my post.

I'd take you up on the challenge to get you debating something else, but the sun came out, and I'm going to go bike riding before it decides to hide for another week.