2020 Canadian Open

[Egidijus Zeromskis]

2. 2020 Bidder - Does anyone know if a bid has yet been submitted for the 2020 Can. Open (Or submitted and accepted)?

There was a bid from Mississauga (see the annual meeting at the CFC forum).

[Don Parakin]

I've always wondered if there was a benefit to accelerated pairings. To see what I mean, consider a tournament with ranked groups Q1 (top), Q2, Q3, Q4 (bottom).

• With un-accelerated, folks in Q1 play folks in Q3 and Q2 plays Q4. Assuming no upsets, Q1 and Q2 will all have a score of 1; the rest score 0. So, for the 2nd round, Q1 plays Q2 and Q3 plays Q4.
• With accelerated, folks in Q1 play folks in Q2 and Q3 plays Q4. Assuming no upsets, Q1 and Q3 will all have 1 score of 1; the rest score 0. So, for the 2nd round, Q1 plays Q3 and Q2 plays Q4.

So, if I'm in Q1, I play someone in Q3 then Q2 with un-accelerated ... or in Q2 then Q3 with accelerated. Not much of a difference. Just reversed.

[Tom O'Donnell]

My understanding is that this is not how (single) accelerated pairings work. Instead, groups Q1 and Q2 get a phantom point before the tournament begins. So in your example, in round two Q1 would only play those also in Q1, since all those players have two points (one phantom, plus one real). Meanwhile Q2 and Q3 would face off. Q2 would all have one (phantom) point, while Q3 would all have one real point. But a point is a point, and they would get paired. Meanwhile all Q4s would play each other since they have no points of either type.

At some point the phantom points get dropped off and the tournament gets paired normally. Of course if it is dropped off for round two that would result in what you wrote, which I think is the point Roger Patterson was making further up the thread.

[Vlad Drkulec]

I prefer sections as otherwise there are too many yoyo pairings.

[Pierre Denommee]

FIDE has become more picky about accelerated pairings. Only accelerated methods approved by FIDE are to be used unless proper authorizations are obtained.

[Bob Armstrong]

Hi Tom:

Thanks for your description of "single accelerated pairing" and how it works.

Roger was right in his suspicion earlier aired that maybe I was not totally clear on the actual working descriptions of the topics I had raised:

single accelerated pairings
double accelerated pairings.

On seeing his post, I decided that it was a problem with this thread that the key terms were undefined. This meant different people could have different definitions, and some people would not know the terms at all. So your clarification saves me some work.

I think it is in Rd. 3, for single accelerated pairings, that the ghost point is taken away - someone can correct me if I am wrong.

Can someone clarify what system Jonathan Berry used in the 2007 Can. Open (Ottawa)?

I thought what he did was to give Q 1 3 ghost points, Q 2 2 ghost points, Q 3 1 ghost point, and Q 4 got no ghost points. Then each successive round for the first three rounds, one ghost point was removed from anyone who had one. Does this system work? I am just now trying to determine what the pairings outcome would be with this, but if others are faster on this, I'd love to hear from them.

It is my personal view that acceleration minimizes (Doesn't totally prevent) the yoyo effect that so many don't like about the one-big swiss. And on the top, the eventual winners have been forced to play much stronger opposition than they do under just normal swiss pairing. Does this observation hold up?

Bob A

[Don Parakin]

Tom, thanks for the correction. Makes more sense. I should've checked with my knowledgeable friend, Wikipedia: LINK.

[Erik Malmsten]

Hi Tom:

Thanks for your description of "single accelerated pairing" and how it works.

Roger was right in his suspicion earlier aired that maybe I was not totally clear on the actual working descriptions of the topics I had raised:

single accelerated pairings
double accelerated pairings.

On seeing his post, I decided that it was a problem with this thread that the key terms were undefined. This meant different people could have different definitions, and some people would not know the terms at all. So your clarification saves me some work.

I think it is in Rd. 3, for single accelerated pairings, that the ghost point is taken away - someone can correct me if I am wrong.

Can someone clarify what system Jonathan Berry used in the 2007 Can. Open (Ottawa)?

I thought what he did was to give Q 1 3 ghost points, Q 2 2 ghost points, Q 3 1 ghost point, and Q 4 got no ghost points. Then each successive round for the first three rounds, one ghost point was removed from anyone who had one. Does this system work? I am just now trying to determine what the pairings outcome would be with this, but if others are faster on this, I'd love to hear from them.

It is my personal view that acceleration minimizes (Doesn't totally prevent) the yoyo effect that so many don't like about the one-big swiss. And on the top, the eventual winners have been forced to play much stronger opposition than they do under just normal swiss pairing. Does this observation hold up?

Bob A

I hate the Swiss yo-yo bounce from master to beginner, better games in playing players within +/- 200 rating points.

The very experienced Jonathan Berry created his own pairing system which the computer couldn't handle. He may have also allowed exceptions so that siblings wouldn't be paired in the early rounds.

Likewise, in Toronto the GTCL tried a quad accelerated pairing divisions which, after ghost points dropped, Swiss Sys barfed up incorrect pairings in the mid-point group (I think highest against lowest rather than middle).

In Cappelle la Grande France, usually 400 entries, they have successfully employed a 3 group division (top third 1 ghost point, middle third 1/2 ghost point) for 3 rounds.

[Roger Patterson]

Hi Tom:

It is my personal view that acceleration minimizes (Doesn't totally prevent) the yoyo effect that so many don't like about the one-big swiss. And on the top, the eventual winners have been forced to play much stronger opposition than they do under just normal swiss pairing. Does this observation hold up?

Bob A

this is not actually true either. Or rather, it depends on where you are on the rating list (and which round). People at the top and at the bottom do play closer in rating - but the rest don't.


As an example, in a regular Swiss pairing after two rounds, the people with 1 point are Q2 and Q3. In an (single) accelerated Swiss the 1 point group is the bottom half of Q1, all of Q2 and Q3 and the top half of Q4. That will clearly result in pairings with a wider range of rating difference for the 1 point group than for the regular Swiss. And for the bottom half of Q1, it's a bigger difference than they would have faced in regular pairings.

[Roger Patterson]

The very experienced Jonathan Berry created his own pairing system which the computer couldn't handle. He may have also allowed exceptions so that siblings wouldn't be paired in the early rounds.

If you are referring to his Kappuskasing system, that experiment was very poorly received and AFAIK, not repeated anywhere. (probably not FIDE approved kosher either)

[Egidijus Zeromskis]

We tried several times accelerated pairings at the club. My impression (meaning that I did not do any big investigations) was that yo-yo was shifted to farther rounds.

[Stephen Wright]

Can someone clarify what system Jonathan Berry used in the 2007 Can. Open (Ottawa)?

Miraculously this link survives: http://cocycc.pbworks.com/w/page/913...Rules#Pairings

The French computer program for this system was unavailable to us, so we did it manually. Three IM norms were achieved - Nikolay Noritsyn, Anton Kovalyov, Bindi Cheng.

[Bob Armstrong]

Hi Stephen:

Thanks for the assistance - Having played in that Can. Open, I know that the manual pairing was a major challenge.

Was it your view, as Victor Plotkin seemed to say above, that this accelerated pairings system worked well - he said it was one of the best tournaments in which he has played?

If so, in what ways do you think your system used was superior to the normal swiss pairing system for large tournaments?

I originated this thread because I want the 2020 Can. Open (Maybe in Mississauga, Ontario?) to be a one-section swiss, with some kind of accelerated pairings system, if we had any input into the bid before finalized and accepted. But I realize there are in Canada some very strong feelings that acceleration does not achieve anything worthwhile in the long run (There are already some such sentiments in this thread).

But I believe in acceleration for long tournaments, as did my recently deceased friend, Phil Haley, and I am trying to see if there could be any majority consensus built favouring this as a Can. Open preference.

Thanks.

Bob A

[Erik Malmsten]

Miraculously this link survives: http://cocycc.pbworks.com/w/page/913...Rules#Pairings

The French computer program for this system was unavailable to us, so we did it manually. Three IM norms were achieved - Nikolay Noritsyn, Anton Kovalyov, Bindi Cheng.

So the ghost points could last a long time. I was wondering about a gradual decrease in the ghost:
For round 1 pairings divided into 5 groups by rating: +2, +1.5, +1, +.5, 0
For round 2 these groups all manually decrease ghost by a half: +1.5, +1, .5, 0, 0
For round 3 decrease again: +1, +.5, 0,0,0
For round 4 decrease again: +.5, 0,0,0,0
For round 5 no more ghost points.

So some bottom group players would get easy pairings for a couple of rounds, but the winners would have to play the middle group before the masters.

[Egidijus Zeromskis]

So the ghost points could last a long time. I was wondering about a gradual decrease in the ghost:

So some bottom group players would get easy pairings for a couple of rounds, but the winners would have to play the middle group before the masters.

A little bit hassle but possible to do this with SwissManager.