Discussion in 'Fencing Discussion' started by gillaspy, Apr 1, 2007.
Let's not bring race into this… :jester:
again, you're not right.
International fencers can count for NIF (the points that add together to make the multiplier) points if they place above enough fencers with NIF points (i.e. within the top 50, more points can be awarded to the international fencer if they finish above certain categories of NIF fencers - 1 pointers, 3 pointers 6 pointers etc)
Gotcha mrbiggs, sorry I misunderstood what you were saying. I do agree that numeric results would probably, at least at the lower levels, be very area-centric as for as results. At the higher levels, people will presumably be attending NACs and large regional events so that they will level out a bit.
Matter of fact, seems to me that one would ONLY need a numeric system for higher level national events, like Div 1 senior NACS and the like. Otherwise you are unlikely to have 20 or 40 A's (ignoring the fact that every rating is effectively 4 based on year) in the room at the same time unless it is some sort of "proto NAC" or "super regional" event like the Cherry Blossom or the Kickoff. Otherwise the system works fine for regular local comps which should be able to be attended for "practice" and to try new things without having to worry about beating everyone 5/15-0.
Nice how I subtly worked in a plug for upcoming awesome local area events eh?
This does bring in one interesting consequence of a numerical system; how do you maintain the weighting of NACs - imagine a fictitious fencer who finishes 3rd in all (4?) mercury cup events but skips div-1 nacs. Should they qualify for summer nationals?
Cville: You worry that people will avoid tournaments in order to prevent their rating decreasing. This is patently unrealistic.
1) In chess, ratings become inactive after a time of playing in no tournaments. You can't just get a ridiculously high rating from a fluke and then expect to be treated as having that rating for the rest of time. So you still have to play in rated tournaments.
2) In chess, nobody avoids tournaments to avoid losing rating points. This is because the inflation that they can achieve from only attending weak tournaments is on the order of tens of rating points, where the range of ratings is from ~700 rank beginner to ~2800 world champion. You can't maintain a particularly large bubble, because if you're significantly overrated, you can attend basically no tournaments without losing rating points, and you have to choose between your rating going inactive and losing points. Most people are not psychotic about losing ten rating points. The big gains in ratings come when you actually improve, and everybody knows they can't game the system for more than a few points.
3) Everybody is worrying about localized bubbles of bad connectivity to the rest of the fencing world. This is actually a much worse problem with the current system than with an Elo system, for the following reason: In our current system, you can really only transmit positive rating changes from division to division, not negative ones. That is, if a division is underrated, you can bring it up to what it should be by going and getting a rating in a national tournament, and then coming back and losing to your clubmates, a common occurrence. That part's fine. The problem is that you can have enormously overrated divisions, and then you can't do anything at all about it, and then an Alaska A is totally different from an NYC A. With Elo, you *can* transmit negative rating changes-- you go to an NAC, lose, come back, win, and everybody's rating gets much closer to the national standard for their level of fencing. This means that if a division has even one person who competes occasionally in national tournaments, and then comes back and competes in local tournaments, that division will be in sync with the rest of the nation. This is not true with the current system.
4) Because of 3), which is to say, a better and more nationally uniform standard for each level of fencing strength, an Elo system would *better* serve to seed tournaments. Everybody tries to excuse the failings of the current system by yelling "It's only used for seeding! It's only used for seeding!" This is fairly irrational. If you need a ranking to seed tournaments, and a new system provides a more accurate, more uniform, and more inflation-resistant way of ranking people, you should switch to the new system.
5) People who are scared of math are missing the point. Elo is really really simple from a fencer's perspective-- if you do better than good people, your rating goes up a lot; if you do better than bad people, your rating goes up a little; if you do worse than bad people, your rating goes down a lot; if you do worse than good people, your rating doesn't go down that much. It's that simple. The difference in two people's ratings is just the best estimate the system has of how likely it is that person A will do better in a tournament than person B. It's not elaborate; it's not complicated. It's nothing like Peter's tournament format suggestion. Many if not most chess players are completely terrible at math and have absolutely no clue how the actual math works, but they find the way their ratings move completely obvious and intuitive, for the above reason.
6)  As you might infer from 5, the Elo system need not pay attention to individual bouts-- it can be modified to take results lists, so that what fencers currently view as important jibes with what the rating system rewards. Paying attention to individual bouts (or worse, individual touches) would totally screw things up, since fencers would place emphasis on different things than they do now, and a rating system need not and should not change that.
I have a question about Elo ratings. Say an A attends a tournament with 25 Us (obviously an unlikely situation). With the current system, he can't lose anything. But say a 2300 shows up to a tournament with 25 300s. His rating can obviously suffer if he loses. He might opt for a black card instead of a loss. With the current black card system, he would be removed from the results, his rating wouldn't suffer, and the ratings of the 300s wouldn't improve much.
How would that be addressed?
Black card = DFL, maybe. That option actually makes more sense in some ways than the current option.
eac, while elo solves the local bubble problem somewhat, it's not a complete solution. That was my point.
George Masin's proposal addressed this situation with the optional designation of some fencers as hors concours. From his proposal:
Incidently, black carded fencers were also listed as excluded from the ratings calculations in the proposal.
The current system has the rather unfortunate problem that U's could be either bad fencers or just fencers who have not recently fenced in USFA competition. So when you mean a tournament with 25 U's and 1 A, you could mean 25 truly bad fencers, or you could mean 25 really good Chinese fencers who for some reason simultaneously decided to show up to one USFA competition.
There is no such confusion with an Elo system. In an Elo system, when you first start competing, you receive what is called a "provisional" rating. This stays for several tournaments, and fluctuates wildly in the first few tournaments. Provisional ratings also mean that they don't affect other people's ratings very much. That is, if you fence someone who's only provisionally rated, your own rating will change by at most 2 or so points, win or lose, no matter the difference between your rating and their provisional rating.
This means that when you see someone who has a non-provisional rating of 700, and you're a 2300 fencer, you are completely assured of kicking their asses in a DE. A 700 is equivalent to someone who's been fencing for less than a year, and a 2300 is equivalent to someone with Div I points-- the person with Div I points would not worry at all, unless he were extremely ill or injured, in which case he should medically withdraw. The reason you have this concern is because of the confusion between provisionally rated fencers and low-rated fencers, which as you now see is unfounded in the Elo system.
eac- What you say holds for pure ELO systems, where each head-to-head match-up is used. Not so much for systems such as the Masin-ELO variant, where the ratings of the field (local field, anyway) and final placement are used. There's no provision for provisionally rated opponents to have a weaker impact. The proposal from 12 years ago specifies that "A fencer of ability who does not have a USFA rating" should be excluded from the ratings calculations, and further text clarifies that this is targetted at foreign fencers.
I don't see why you couldn't have a provisional rating not affect the other people's ratings. You could just say that for a person's first 5 (to pick a random number) tournaments, it was for the purposes of other people's rating changes as if they weren't there (so, in particular, everybody who placed below them was as if they placed one place up). This would work fine for determining someone's rating fairly accurately before having their results affect others' ratings, and I don't see why it should be restricted only to "fencers of ability." A beginning fencer should have the same thing-- everybody who has no rating or a very old, inactive rating should act this way. During the first 5 tournaments, the rating could fluctuate wildly, just like a provisional rating in chess.
Ugh! Not this again.
I don't know why you say ELO
I say goodbye!
That would have gotten points for creativity, except for the misssspellling.
Goddness gracious, you're a picky one. :mutant:
In the British system, the points are calculated by a national body. Therefore, local people do not have to worry about anything else than sending in the results list. No calculation for them to do whatsoever.
Have a nice time!
Peter, for most Americans, we've heard this line before, and it's not very reassuring....
I think one problem may very well be one of scale. If I understand it right, the BFA Senior Mens Foil points list has a little over 500 people. That is presumably for the entire country. If I am wrong on that someone please correct me. My division (and there are more divisions than there are states in the US...) also has a rolling points list. We have 169 mens foilists and my division is far from a fencing powerhouse as far as size goes. The USFA has something like 22000 active competitive members, many of who fence multiple weapons, who's scores would have to be tracked, recorded and updated in a timely manner. By a national office that still demands things be faxed or mailed into them... Not the most confidence inspiring as Allen mentions.
I am not bulleting these off to be argumentative, I just have several responses/questions and if I try to type them straight, I am afraid they will loose their context or not make sense...
How so? Right now most tourneys for me that are not NAC's, huge super regional events or qualifiers are practice tourneys. I can go there, try new things and not worry about my standings except on our local points list. I don't see how that could be the same under an ELO system. Especially since many of the ones I have seen count each touch or each pool bout as a win/lose! Also people who were chasing the points total for team placement and such, I think they would determine what events would give the most points for the least effort/risk (like some do with NAC's or FIE events) and avoid the rest. Further I am worried that it would hurt local attendance. I already see a some of the "I already have an A or B, why go out to a local tourney and give them practice against me" mindset.
And the same with the current system. A letter degrades every year if you don't renew it. After 4 years it drops to the next letter, as you know very well. I would like to see the cap be 2 years instead of 4, but it seems that this point is already served by the current system.
Not to put to fine a point on it, but Chess is a board game. It is not a sport. Fencing is a sport with all the injuries, conditioning peaks and so forth that go with a sport. I can play chess with a pulled hamstring and a hangover. I could fence with a pulled hamstring and a hangover also (hell, I do...). The difference between now and an ELO future is that if I lost to a scrub because my leg gave out, I would not loose my rating, or have it lowered. Physical factors, injuries and peak times in training cycles as well as stress from life/school events all seem to point to people skipping more events under an ELO system, IE only fencing in events where they are at a strong time in their training/development.
I agree that it is a problem under the current system, but again it is mainly a problem at the very top levels of fencing. On a local level, IE the vast majority of US fencing, seedings work out about where the should. I think we would all be better served by something simple (just spit balling here...) like expanding the points list to 100 for each weapon and decree that national points must be factored into seeding for all events. Between the degrading of ratings over time, the inclusion of expanded national points and the system as it stands, I think the system would be more accurate than now without all the chaos and problems of an ELO type system.
I think you may be missing the point. Or at least the difference between a board game and a sport. Or maybe I am still missing some key point of the ELO system. Regardless, scared of math or not the fact remains that it would seem rather difficult to have this done in a transparent fashion to assure that things were done correctly at a local and national level and to insure that someone in the USFA of Division were not say, skewing the results so that there kids/students would make the national team. Now it is pretty easy, even my math skills are up to it. If it is done by some archaically coded database that only the USFA has access to, that means that we have to have total faith in the USFA. For me, that would require overhauling a hell of a lot more than just the ratings system!
I think I would love to see a mock up system that took those sorts of things into account. I have not seen that so far and that might be part of the reason for my resistants.
Again, no trying to argue. You seem to have a pretty good grasp of the ELO and are a fan of it. I am trying to understand your point of view better. They I might start arguing!
Now, although at first this sounds like a bad idea due to the she shear size difference, this would actually be pretty easy to all calculate automatically using a pc. I can see it integrated very well into FRED for example: each organiser enters results, FRED automatically calculates their total amount of points, and adds them to the ranking. With such a system, it should be pretty easy to determine both divisional and national rankings, with very little hastle.
Just a thought.
Have you played competitive chess?
I play very low level tournament chess, but my game is certainly affected by my by both my physical and mental condition and conditioning. You should read about what some chess grand masters do to stay in shape. It is very difficult to concentrate consistently over a 2 hour chess game when everything feels right. Lose your concentration and you lose the game--not so different from fencing, in my experience.
I don't understand the objection to using individual matches in the ELO calculation.
Separate names with a comma.