The individual fencing statistic

["Fence like a Cat"]

The Individual Fencing Statistic

By. Marty Tetloff
9/26/2004

In doing some thinking and research for a personal project on training performance I have begun to consider the bigger picture of the individual fencing statistic and it’s meaning and use for the greater good of fencing. I am more concerned about the how and why we maintain statistics at this point and how can or would that provide some value to the sport in the longer term and bigger picture. Somewhere I read a criticism about fencing in it’s lack of statistics except for the national level results and rankings.

In my research I have found there is certainly value in maintaining win loss records and bout details in periodization planning as an evaluation tool. This provides for a level of detail to include ‘tournament level “sharp” bouts fenced according to Aladar Kogler in his book Planning to Win, with victories versus defeats, touches scored versus touched received as one measure of intensity and performance level. I could see in tournaments the actual placement and rankings earned as having value in this regard as well when put into this context.

Some would consider the baseball fan who pours over the daily sports page tabulations crazed in their capacity to cite this or that statistic about a favorite ball player. We monitor and watch Barry Bonds as he surpasses 700 homers. Then there is Ichiros battle to break the base on balls hit record. The list of statistics is virtually endless and each sport has it’s own unique set in maintaining a relatively detailed level of the individual performance statistic. These stats have been maintained for as long a baseball has been around. Fencing has been around a very long time indeed. Why doesn’t fencing have this level of detail and legacy? Is it because it’s an individual sport at it’s basic nature? I believe we have the capacity to invent a common set of statistics that we can promote based on what data we collect and report. I would also submit that this information over time would have true value for the greater good of the sport.

In fencing I note in various college fencing media guides cumulative win loss records of individuals by year, weapon and wonder about the ramifications and actual use of these statistics. The NCAA polls rank colleges at least twice a year but that’s based on college or division groupings presumably but I don’t actually know. How are they used? I know if I were looking for a fencing program for college I would certainly examine their results.

Then there are the points awarded in the various qualifying paths in the United States Fencing Association. Do these numbers have a greater meaning then who qualifies? Why isn’t there some book out there that lists every fencer in the United States Fencing Association’s history with some common baseline statistic?

Somewhere along the line in my own fencing career I’ve wondered can I actually get my hands on all those tournament results reports to the USFA. I certainly haven’t kept results. But these would only include actual tournament final rankings I would think and this or that classification earned or re-earned. The pool results are probably lost to the ages unless there is someone out there in the various places I’ve lived that considers archival important. I will also tell you that my career hasn’t been that remarkable, but it is my fencing career. I’ve been a fan of the Canadian Fencing Federation’s Passport; the little red book that has an individual’s long term effective fencing career at their own fingertips.

I was told by a reliable source that Golubitsky can cite his own performance statistics long term at a level of detail that would impress the average ballplayer stat jocky.

Now the question is, what would the value be to the individual and sport itself if long term detailed level longitudinal statistics were maintained for an individual fencer? Certainly it would depend on what stats were maintained. But off the cuff I suppose events fenced, win loss record, etc. with a time element would show overall performance.

Statistics like these though by themselves have limited value except for the individual and their coach in evaluating a training program. To have real meaning and value in the sport there must be comparisons made. This assumes some common formula and standard.

We have a few that can be cited immediately:
Ranking changes
Tournaments fenced, number
Pool and DE results
Wins, cumulative
Defeats, cumulative
Touches scored, cumulative
Touched received, cumulative
Success percentage
Potential subset by left or right handed competitor

I would think some kind of ratio would be good, like baseballs earned run average (ERA). Currently some schools report percentage of win numbers.

The Western Washington Division has annual weapon perpetual awards based on the performance in four open events. Points are awarded based on placement typically to 32nd place. The individual with the greatest points per weapon earns the title. Cumulative totals for the three weapons determine the three weapon champion.

To some end I am formulating a program for my club fencers based on the relatively discreet elements noted above. Each fencer who attends a USFA event makes sure they get a copy of their pool forms and the detailed DE event results. I plug them into a spread sheet and maintain their basic statistics. I will do this for the season and see what happens.

Our club also has an internal competitive ladder and we award points toward the quarterly award of medals based on their activity. We also have small internal competitions and have recently come up with a scheme to assign points based on results for a perpetual trophy based on weapons fenced including three weapon activity recognition.

Now how does this help in my thesis? Without source documents it would be difficult to actually construct reliable statistics. What I will maintain is a relatively small fencer population for which I will make some comparisons. This information will certainly be used for performance analysis.

My question to you then is how do we get the ball rolling and do we want to? Is there value in maintaining and reporting individual statistics outside of the training regime and would that provide meaning and context across the sport if such things were done?

Lastly, I remember some discussion somewhere that talked to the value of rankings and how these don’t really show the actual skill and participation over time. This could be one measure that would add to the dialog and potentially point toward a new set of rankings that factors in “time in” the sport so to speak. Wouldn't it be interesting to know how seasoned a fencer is and how they are fencing currently or even in the past based on some measure of performance?

I would be interested in any feedback on the concepts included in this brief paper and would consider a longer essay for publication based on your thoughts.

To reduce the amount of traffic in the newsgroup I would suggest that your notes be sent to my email address directly at tacoma_fencing@yahoo.com. Feel free to discuss this with the group of course.

Thanks for your thoughts.

Marty Tetloff

[swordsen]

Interesting to say the least.....

[achilleus]

There's an epee fencer I know who would love stats.
He takes it to an extreme, but it makes sense. Some of the stats he talks about are:

Win/Loss
Head to Head (i.e. When Golubitsky and Gregory drew each other, the announcers could discuss the head to head)
Bout length coupled with win/loss (for example what percentage of the total wins finished in the 1st period? The 2nd? The 3rd? How about losses?)
Point differential (does the fencer win bouts by a large margin? Or are they all 15-14? Does are fencer tend to lose close bouts?)
Overtime victories with and without priority
For foil and saber, what % of the touches are scored as the attack? Riposte? Counter?

Those are the few I remember...

I think he also mentioned preferred target areas...

[D+F+P=Hadouken!]

There's an epee fencer I know who would love stats.
He takes it to an extreme, but it makes sense. Some of the stats he talks about are:

Win/Loss
Head to Head (i.e. When Golubitsky and Gregory drew each other, the announcers could discuss the head to head)
Bout length coupled with win/loss (for example what percentage of the total wins finished in the 1st period? The 2nd? The 3rd? How about losses?)
Point differential (does the fencer win bouts by a large margin? Or are they all 15-14? Does are fencer tend to lose close bouts?)
Overtime victories with and without priority
For foil and saber, what % of the touches are scored as the attack? Riposte? Counter?

Those are the few I remember...

I think he also mentioned preferred target areas...

Does he also publish a book or anything like that, so I can see these statistics?

[SwordSoul]

ya know, i've wondered the same thing. I am going to start implementing W/L tables and results, etc... logs, but i was also going to incorporate aspects of nutrition, training for that week, sleep, mood, mindset, warm-up routine before the competition, time spent AT the tournament vs. actual strip time, mood after the bout... although for example if you're trying to get a scholarship or some such thing, they won't be altogether interested what you ate or how much sleep you got before your competitions-- or would they? I recently read an article about NHL coaches who select their team based on their effort, not really their ability. Not that ability isn't very very important, and it IS an individual sport here...

so, one effect of something like this i can see an overall "database" on statistics is "wow fencer A has an x/y hit to recieved ratio!" or it could subtly change the way some people train: i.e. if others will see how many hits they recieve even if their record is good it may prompt them to spend more time practicing defense?? i dunno. I think a video camera or someone monitoring your bout (Everyone seems to know their "flick success" stat!) could be good for reference.


Does anyone know how this concept is applied in say, boxing?

matt

[Switch]

The Individual Fencing Statistic

By. Marty Tetloff
9/26/2004

In doing some thinking and research for a personal project on training performance I have begun to consider the bigger picture of the individual fencing statistic and it’s meaning and use for the greater good of fencing. I am more concerned about the how and why we maintain statistics at this point and how can or would that provide some value to the sport in the longer term and bigger picture. Somewhere I read a criticism about fencing in it’s lack of statistics except for the national level results and rankings.

In my research I have found there is certainly value in maintaining win loss records and bout details in periodization planning as an evaluation tool. This provides for a level of detail to include ‘tournament level “sharp” bouts fenced according to Aladar Kogler in his book Planning to Win, with victories versus defeats, touches scored versus touched received as one measure of intensity and performance level. I could see in tournaments the actual placement and rankings earned as having value in this regard as well when put into this context.

Some would consider the baseball fan who pours over the daily sports page tabulations crazed in their capacity to cite this or that statistic about a favorite ball player. We monitor and watch Barry Bonds as he surpasses 700 homers. Then there is Ichiros battle to break the base on balls hit record. The list of statistics is virtually endless and each sport has it’s own unique set in maintaining a relatively detailed level of the individual performance statistic. These stats have been maintained for as long a baseball has been around. Fencing has been around a very long time indeed. Why doesn’t fencing have this level of detail and legacy? Is it because it’s an individual sport at it’s basic nature? I believe we have the capacity to invent a common set of statistics that we can promote based on what data we collect and report. I would also submit that this information over time would have true value for the greater good of the sport.

In fencing I note in various college fencing media guides cumulative win loss records of individuals by year, weapon and wonder about the ramifications and actual use of these statistics. The NCAA polls rank colleges at least twice a year but that’s based on college or division groupings presumably but I don’t actually know. How are they used? I know if I were looking for a fencing program for college I would certainly examine their results.

Then there are the points awarded in the various qualifying paths in the United States Fencing Association. Do these numbers have a greater meaning then who qualifies? Why isn’t there some book out there that lists every fencer in the United States Fencing Association’s history with some common baseline statistic?

Somewhere along the line in my own fencing career I’ve wondered can I actually get my hands on all those tournament results reports to the USFA. I certainly haven’t kept results. But these would only include actual tournament final rankings I would think and this or that classification earned or re-earned. The pool results are probably lost to the ages unless there is someone out there in the various places I’ve lived that considers archival important. I will also tell you that my career hasn’t been that remarkable, but it is my fencing career. I’ve been a fan of the Canadian Fencing Federation’s Passport; the little red book that has an individual’s long term effective fencing career at their own fingertips.

I was told by a reliable source that Golubitsky can cite his own performance statistics long term at a level of detail that would impress the average ballplayer stat jocky.

Now the question is, what would the value be to the individual and sport itself if long term detailed level longitudinal statistics were maintained for an individual fencer? Certainly it would depend on what stats were maintained. But off the cuff I suppose events fenced, win loss record, etc. with a time element would show overall performance.

Statistics like these though by themselves have limited value except for the individual and their coach in evaluating a training program. To have real meaning and value in the sport there must be comparisons made. This assumes some common formula and standard.

We have a few that can be cited immediately:
Ranking changes
Tournaments fenced, number
Pool and DE results
Wins, cumulative
Defeats, cumulative
Touches scored, cumulative
Touched received, cumulative
Success percentage
Potential subset by left or right handed competitor

I would think some kind of ratio would be good, like baseballs earned run average (ERA). Currently some schools report percentage of win numbers.

The Western Washington Division has annual weapon perpetual awards based on the performance in four open events. Points are awarded based on placement typically to 32nd place. The individual with the greatest points per weapon earns the title. Cumulative totals for the three weapons determine the three weapon champion.

To some end I am formulating a program for my club fencers based on the relatively discreet elements noted above. Each fencer who attends a USFA event makes sure they get a copy of their pool forms and the detailed DE event results. I plug them into a spread sheet and maintain their basic statistics. I will do this for the season and see what happens.

Our club also has an internal competitive ladder and we award points toward the quarterly award of medals based on their activity. We also have small internal competitions and have recently come up with a scheme to assign points based on results for a perpetual trophy based on weapons fenced including three weapon activity recognition.

Now how does this help in my thesis? Without source documents it would be difficult to actually construct reliable statistics. What I will maintain is a relatively small fencer population for which I will make some comparisons. This information will certainly be used for performance analysis.

My question to you then is how do we get the ball rolling and do we want to? Is there value in maintaining and reporting individual statistics outside of the training regime and would that provide meaning and context across the sport if such things were done?

Lastly, I remember some discussion somewhere that talked to the value of rankings and how these don’t really show the actual skill and participation over time. This could be one measure that would add to the dialog and potentially point toward a new set of rankings that factors in “time in” the sport so to speak. Wouldn't it be interesting to know how seasoned a fencer is and how they are fencing currently or even in the past based on some measure of performance?

I would be interested in any feedback on the concepts included in this brief paper and would consider a longer essay for publication based on your thoughts.

To reduce the amount of traffic in the newsgroup I would suggest that your notes be sent to my email address directly at tacoma_fencing@yahoo.com. Feel free to discuss this with the group of course.

Thanks for your thoughts.

Marty Tetloff

I just felt to make another really long reply. haha!!!

[JackSparrow]

I like the Vic Reeves quote,
"67% of statistics are made up on the spot"

["Fence like a Cat"]

I am trying to lay my hands on an AFLA publication issued in 1971 titled: Fencing Record Book. It was basically a competitive performance results journal and was issued to annual and life members as announced in the Nov/Dec 1971 issue of American Fencing.

A photocopy of a blank or used copy would be sufficient. I would be happy to pay any cost involved in reproduction or mailing.

Thanks

Marty Tetloff

[Goldgar]

Some of the stats he talks about are:
[...]
Head to Head (i.e. When Golubitsky and Gregory drew each other, the announcers could discuss the head to head)

I was impressed to discover, while looking up a clubmate's results on the FIE website the other day, that they have these head-to-head results tallied, for the World Cup events at least. Anyone can look up a fencer and see how that fencer fared in all their WC encounters with another fencer. That's cool.

[ysbadadden]

I think that solid and objective fencing stats are feasible.
Once again: we first need a top-to-bottom points system, like chess. These systems exist and are used in fencing -- just not here. The letter ranking system we have is inadequate.
We need the head-to-head results already mentioned. The touches scored/received ratio would be a good supplement to the W/L.
If I were team statistician, I'd want to know time of bout (already mentioned) but also the time and order of the touches.
And I'd want to know how the touches were scored: on attacks, counter-attacks, ripostes, etc. I wouldn't need video replay for this; I could get it from the ref's call.

Then we could say of someone: He's got a 17-3 W-L, and only a .555 touch average, but it's .711 in the third period -- 12 of his wins have come from behind in the third period. In the first period, he scores 45% of his touches on ripostes, but his attacks gain power and frequency as the bout goes on.

These are all statistics that are available to anyone scoring at the match -- you wouldn't have to be an expert. I'm not sure that you could come up with a stat like "provoked opponent's attack from bad distance," etc.

["Fence like a Cat"]

I created a Facebook group on statistics and am developing a model statistical manual for fencing.

http://www.facebook.com/group.php?gid=79755251751

marty

[noodle]

well, for what its worth, since its actually making progress, i'm developing an online version of a fencing journal specifically designed with statistics and analytics gathering as its strength.

[tkexi991]

There's an epee fencer I know who would love stats.
He takes it to an extreme, but it makes sense. Some of the stats he talks about are:

Win/Loss
Head to Head (i.e. When Golubitsky and Gregory drew each other, the announcers could discuss the head to head)
Bout length coupled with win/loss (for example what percentage of the total wins finished in the 1st period? The 2nd? The 3rd? How about losses?)
Point differential (does the fencer win bouts by a large margin? Or are they all 15-14? Does are fencer tend to lose close bouts?)
Overtime victories with and without priority
For foil and saber, what % of the touches are scored as the attack? Riposte? Counter?

Those are the few I remember...

I think he also mentioned preferred target areas...

Hey, method of scoring is important for epee too! Simple attack, parry-riposte, counterattack and double vs. single lights are all important to track.

[Sukinator]

This is an area where askFRED can make a big contribution!

[Displacement]

Something like the after round boxing statistics would be cool (but impracticle in all but the highest level events). Instead of punches thrown, punches landed, jabs, uppercuts etc. you can have attacks made, attacks landed, parry repost, counter attacks. Pretty interresting idea.

My college team has and continues to keep very detailed records (back to the early 40s). I can go back and find out every bought I fenced, what team it was against and if I won or lost. No Names or scores however.

["Fence like a Cat"]

... My college team has and continues to keep very detailed records (back to the early 40s). I can go back and find out every bought I fenced, what team it was against and if I won or lost. No Names or scores however.

Good stuff Displacement. I too have recorded such stats since college in my somewhat non illustrious career...but I also recorded indicators, event, placement, location and level as well.

["Fence like a Cat"]

Good relevant quotes from Frederick Mosteller, in Anthology of Statistics in Sports;: "Lessons from Sports Statistics"
"Lesson 1. If many reviewers are both knowledgeable about the materials and interested in the findings, they will drive the author crazy with the volume, perceptiveness, and relevance of their suggestions..."
"Lesson 2: I you develop some inferential statistical methods, you are likely to have a use for them in a paper on sports statistics."
"Lesson 3: There is always more to do."
"Lesson 12: Getting additional people involved in statistical work is a beneficial activity, and they should not have to recruit themselves. Can we do more of this?"

"My experience with news writers in relation to sports statistics strongly suggests to me that statisticians should do more about getting statistical material intended for the general public in a more digestable fashion."

[teacup]

I was impressed to discover, while looking up a clubmate's results on the FIE website the other day, that they have these head-to-head results tallied, for the World Cup events at least. Anyone can look up a fencer and see how that fencer fared in all their WC encounters with another fencer. That's cool.

Besides the FIE website, the German website for live results has a great deal of information about every tournament and a fencer's historic results.

For example: Click on a World Cup, reports for tournament results include all pools, DE's with touch by touch breakdowns of the bouts.

click on a fencer to see their photo, personal information and statistics on all the bouts they have fenced.

http://www.nahouw.net/ also has detailed statistics about fencers including graphs of their performance over the years.

The USFA does scan the pool sheets and DE's eventually but prior season's results are no longer on the USFA website.

[Inquartata]

Oh ye gods, save us from the math wonks!

I will fight this sort of thing with every breath in my body...and I haven't got that many of those left!