Complicating Canadian Rankings

[kalivor]

The CFF has just posted a new system for rankings, where points for a tournament are earned as follows:

R = F * (1.006 – (log P/log N))
where:
R = ranking points awarded for P
P = place achieved in the competition
N = number of entries in the competition: and
F = 15xA + 10xB + 5xC + 3xD
Where A = # As present at the competition
B = # of Bs present at the competition
C = # of Cs present at the competition
D = # of Ds present at the competition
A minimum force (F) will be applied for all sanctioned competitions.

(whole thing here: http://www.fencing.ca/downloads/rule...g_2006_eng.pdf)

Thoughts? Opinions?

Personally, I think it's a bit over-complex. I rather liked the Canadian system before: Every fencer had a point value (A=41(?), B=21, C=11, D=6, U=1), and the winner of the tournament got the sum of all the fencers' points (plus a first place bonus), the second place person got the sum of all the fencers' points (minus those for the top rated fencer), plus a second-place bonus, and so on until the last place person got one lonely point.

Perhaps the new system will be more accurate in determining an overall ranking, but the old system really didn't seem bad to me. It was rather transparant, and made it fairly easy to gauge how many points would be gained for a given tournament and placing.

(The CFF has also introduced a bunch of other changes, for which I either have not read them yet (mostly) or don't feel are worthy of discussing -- they should be checked out, particularly by Canadian fencers: http://www.fencing.ca)

So ... what do people think of the changes? Good? Bad? At the very least, it looks like membership fees are dropping.

[KD5MDK]

This sounds a lot closer to the system the US uses for awarding domestic points for FIE events. However, because I don't know how the A-D system in your country works, I can't say if it sounds good to me.

[Inquartata]

Thoughts? Opinions?

I can muster only "Ack! Bleargh!"

[KD5MDK]

I can muster only "Ack! Bleargh!"

It was worth a try, but even though your choosing not to include a poll kept him from saying "flawed", we still haven't managed to strike him entirely speachless.

[jBirch]

The CFF has just posted a new system for rankings, where points for a tournament are earned as follows:

R = F * (1.006 – (log P/log N))
where:
R = ranking points awarded for P
P = place achieved in the competition
N = number of entries in the competition: and
F = 15xA + 10xB + 5xC + 3xD
Where A = # As present at the competition
B = # of Bs present at the competition
C = # of Cs present at the competition
D = # of Ds present at the competition
A minimum force (F) will be applied for all sanctioned competitions.

(whole thing here: http://www.fencing.ca/downloads/rule...g_2006_eng.pdf)

Thoughts? Opinions?

Personally, I think it's a bit over-complex. I rather liked the Canadian system before: Every fencer had a point value (A=41(?), B=21, C=11, D=6, U=1), and the winner of the tournament got the sum of all the fencers' points (plus a first place bonus), the second place person got the sum of all the fencers' points (minus those for the top rated fencer), plus a second-place bonus, and so on until the last place person got one lonely point.

Perhaps the new system will be more accurate in determining an overall ranking, but the old system really didn't seem bad to me. It was rather transparant, and made it fairly easy to gauge how many points would be gained for a given tournament and placing.

(The CFF has also introduced a bunch of other changes, for which I either have not read them yet (mostly) or don't feel are worthy of discussing -- they should be checked out, particularly by Canadian fencers: http://www.fencing.ca)

So ... what do people think of the changes? Good? Bad? At the very least, it looks like membership fees are dropping.

I got a copy of this proposal from the AFA a couple of weeks ago and forwarded my thoughts on it to them. I don't like it.

My objections were threefold:

1) There is a drastic point difference between 1st and 2nd place in any given tournament.

2) The only thing that really matters for points is the number of A's at the tournament. Everyone else is just fodder.

3) This proposal is concurrent with the membership structure change. If both go through, all the existing A's (which are concentrated in Ontario and Quebec) will generate A's, B's and C's at the tournaments they frequent, essentially killing any tournament outside of those regions. Up and comers, who need national points, will flock to tournaments in central Canada at the detriment of tournaments in the west and the east. I don't like the idea of concentrating all the good fencers in one region and think that doing so artificially through the point system is a bad idea. It's bad because the problem with fencing in Canada isn't identifying high performance athletes: it's getting them to join in the first place.

It's a good idea that the CFF wants to make the quality of a competition matter, they just screwed up the math.

James.

[epeeisky]

When you need more than a simple calculator to figure out your points there is something wrong.

[Fencergrl]

They can make it as complicated as they like... but I agree with JBirch on this. I don't like a system that makes it nearly impossible to gain a rating anywhere but Ontario or Quebec.

[keropie]

It definitely seems kinda retarded. 2nd place in a 100 person event gets like 86% of the points as first place? That does seem harsh. And it sure does take a while to figure out points (though I'm sure it'll be porgrammed into something, of course), and it just seems fairly unnecessary. Any idea what the reasoning behind all of this was, or who came up with the formula and such?

[jBirch]

Not sure who came up with the formula, but it's been applied in other sports with prior rankings to determine points with a large leaning towards quality over quantity.

The idea is that the quality of the competitors matters more then the quantity of the competitors. There's been some complaints over the past few years that huge regional competitions are "inflating" the rankings of some of the competitors and skewing the results. The argument is that if you come in first in a 100 person tournament with all scrubs you get more points then if you come in first in a tournment with 6 people, 3 of which are A's. Shouldn't you get more points for beating the 3 A's then if you beat 99 beginners?

The biggest problem I have is that the current rankings are based on the Top X % in a given weapon in the country and they are concentrated in Toronto and Montreal. At the same time as this change is proposed, the CFF is restructuring how they classify "members" so next year you'll have a huge influx of unrated fencers. What this creates is a "points vacuum" around the current A's and B's where any tournament they go to in the next year will promote a whole crop of these unrated fencers by mere virtue of their proximity. So those tournaments that have A's will have a huge amount of points up for grabs...which means that those fencers that go to them will get more points...which will make them more important to go to for everyone (including the current A's)...which means that wherever the current A's are clustered, that's where all the good fencing needs to be. This leads to a concentration of meaningful fencing in just a handful of communities accross the country.

It also makes the survival of the provincial circuits dependant upon their ability to recruit (not train) the current A's.

James.

[Wise-Epeeist]

That's unnecessarily complicated. But then Canada rarely conforms to any of the standards recognized by the rest of the states in the USA. Canadians think they're special or something.

[Fencergrl]

Yeah... those Canadians actually think they are a different country

[KD5MDK]

Yeah... those Canadians actually think they are a different country

Isn't it cute? We even let them have our political party.*


(I'm speaking of the CRAP, of course)

[Fencergrl]

Okay... I'll bite, what's CRAP?

[KD5MDK]

Canadian Conservative Reform Alliance (Party)

[glowstix]

I don't like a system that makes it nearly impossible to gain a rating anywhere but Ontario or Quebec.

...............well i can look at the bright side: i have access to the best epee coach in the country now!!

[kalivor]

I got a copy of this proposal from the AFA a couple of weeks ago and forwarded my thoughts on it to them. I don't like it.

My objections were threefold:

1) There is a drastic point difference between 1st and 2nd place in any given tournament.

I don't see this. They give two examples in the pdf and, assuming that their calculations are correct, there's less difference. In their example with 3As, 5Bs, 9Cs, 4Ds (which IMO is a fairly strong open level competition), the difference between first and second is only 28 points. In the old system, with an A in the mix, the difference would have been larger.

In the example without As, Bs or Cs, the difference is also smaller than under the old system (5.7 points vs. more than 10 with a D in the mix).

If anything, the difference between first and second has been made smaller.

2) The only thing that really matters for points is the number of A's at the tournament. Everyone else is just fodder.

I'm not sure about this. The criteria for getting an A means that there are very few of them in the country (MF, for instance, has three ... one of whom isn't likely to show up at anything below a CSC level). It seems very likely that the number of Bs, Cs, OR Ds will be sufficient to be worth as many (if not more) points as the As are, at most competitions.

3) This proposal is concurrent with the membership structure change. If both go through, all the existing A's (which are concentrated in Ontario and Quebec) will generate A's, B's and C's at the tournaments they frequent, essentially killing any tournament outside of those regions. Up and comers, who need national points, will flock to tournaments in central Canada at the detriment of tournaments in the west and the east. I don't like the idea of concentrating all the good fencers in one region and think that doing so artificially through the point system is a bad idea. It's bad because the problem with fencing in Canada isn't identifying high performance athletes: it's getting them to join in the first place.

Here I agree, though I feel that the old system did this as well. Though I'd like to note that domestic tournaments can't generate As in Canada.

[samh]

Glowstix I'm curious who you're calling the best epee coach in the country? Are you training with Leslie right now?

[Fencergrl]

Glow's moved to "God's Country".... Calgary, Alberta (okay we won't tell you what part of the body that Calgary is referred to )

He’s still adjusting to the word “loonie”.

[RoninX]

They can make it as complicated as they like... but I agree with JBirch on this. I don't like a system that makes it nearly impossible to gain a rating anywhere but Ontario or Quebec.

Is it currently a reasonable possibility? There are many areas in the US where gaining an A or B is virtually impossible when attending local/regional tourneys.

[Fencergrl]

It is very difficult to gain any sort of a rating in Canada as it is. I don't want to see it regionalized further.