poll by TD: predefining the sections

[Aris Marghetis]

The RA Fall Open has just been wrapped up, and my next weekender is the National Capital Open, scheduled for the beginning of December. In preparation, I have been mulling over how I structure my sections, largely depending on how many people have registered in each rating range. The advantage of such an approach is that, at the last minute, I can customize the section cutoffs, especially to facilitate reasonable pairings for 5 rounds.

The disadvantage is that especially for players in the 1800 range, they could either be paired mostly up in a section for 1800 and over, paired mostly down in an U1800 section, or if there are more players than usual, paired evenly in a section from 1600 to 2000.

Therefore, I am wondering if it would be a better idea to guarantee a minimum of 3 sections: >=2000, U2000, U1600. That is the poll question I am attaching to this thread.

A couple of closing notes :

- In tracking my tournaments, the number of players above 2000 and above 1600 have been consistent. In other words, only a small U1600 section "might" not make for the best pairings. However, I feel that it would work to pseudo round-robin a smallish U1600 section.

- If any of these 3 sections approach 2 dozen players, then I would consider splitting that big section into 2 smaller sections, like splitting 1600-1999 into 1600-1799 & 1800-1999.

Thanks for your input, it is much appreciated, Aris.

[Aris Marghetis]

Re: poll by TD: predefining the sections

The RA Fall Open has just been wrapped up, and my next weekender is the National Capital Open, scheduled for the beginning of December. In preparation, I have been mulling over how I structure my sections, largely depending on how many people have registered in each rating range. The advantage of such an approach is that, at the last minute, I can customize the section cutoffs, especially to facilitate reasonable pairings for 5 rounds.

The disadvantage is that especially for players in the 1800 range, they could either be paired mostly up in a section for 1800 and over, paired mostly down in an U1800 section, or if there are more players than usual, paired evenly in a section from 1600 to 2000.

Therefore, I am wondering if it would be a better idea to guarantee a minimum of 3 sections: >=2000, U2000, U1600. That is the poll question I am attaching to this thread.

A couple of closing notes :

- In tracking my tournaments, the number of players above 2000 and above 1600 have been consistent. In other words, only a small U1600 section "might" not make for the best pairings. However, I feel that it would work to pseudo round-robin a smallish U1600 section.

- If any of these 3 sections approach 2 dozen players, then I would consider splitting that big section into 2 smaller sections, like splitting 1600-1999 into 1600-1799 & 1800-1999.

Thanks for your input, it is much appreciated, Aris.

I just thought of something, which is especially with the second option of 3 predefined sections, there happen to be players who straddle ratings of 2000 or 1600, often crossing over one way or the other. This is to the point that sometimes, right before a tournament, a previous tournament they played in just got rated, and it changes the section that they play in, which might have affected their decision to even register.

Therefore, especially if you vote for the second option of 3 predefined sections, I would like to know what you think of placing players in the section of their peak CFC Regular rating, as opposed to the highest of their FIDE and CFC ratings, which is what I do now.

Thanks again, and best regards, Aris.

[Felix Dumont]

Re : poll by TD: predefining the sections

I don't know if it's the same in your tournaments, but in Quebec the sections which have the more players are frequently the U1400 or U1200, so if I was you I would do at least 4 sections.

But 3 sections is clearly better than no predefined sections at all.

[Aris Marghetis]

Re: Re : poll by TD: predefining the sections

I don't know if it's the same in your tournaments, but in Quebec the sections which have the more players are frequently the U1400 or U1200, so if I was you I would do at least 4 sections.

But 3 sections is clearly better than no predefined sections at all.

That's interesting about the lower rating groups in Quebec. In my events, there is a slow but definite trend upwards, with U2000 usually being my biggest section. My U1600 and U1400 are slowly but surely shrinking, and I've never even considered any U1200.

By the way, we get players to come to Ottawa from Toronto, Kitchener, and even further ... but what would it take players like yourself from Montreal to come more?

Thanks for your feedback and vote.

[Jason Lohner]

Re: poll by TD: predefining the sections

Definitely in a larger tournament predefined sections are nice. I would also state that those people rated under 1600 are average 'club' players... these people now have to play 2x the fee to play now for a yearly tournament. Don't be surprised if people of this strength show up in fewer numbers now.

[Caesar Posylek]

Re: poll by TD: predefining the sections

The second choice will be the certain majority of votes. No doubts in this topic.
:)

[Aris Marghetis]

Re: poll by TD: predefining the sections

The second choice will be the certain majority of votes. No doubts in this topic.
:)

Yes, so far, the voting is overwhelmingly for predefined sections. Unless there is a dramatic swing in the voting, I humbly (and gratefully) will change my section format.

Now, any thoughts on the more far-out idea of using peak ratings to determine what section a player gets paired in? I think the cons outweigh the pros. What do you think?

[Bob Armstrong]

Peak Rating Qualification?

Hi Aris:

I believe peak ratings is a good concept. It does discourage sandbagging. It is discouraging to U 2000 players when an expert, temporarily off his game in the prior one or two tournaments, has just slipped below 2000, and decides to play in the U 2000. He may be too strong for the section, and kicking him upstairs would seem to be fair enough.

I think it should be the peak rating in the last 2 years ( in any of the ratings systems ). I like the 2 prior years span, but I suppose it could be lessened to the prior year.

Just my 2 cents worth.

Bob

[Aris Marghetis]

Re: Peak Rating Qualification?

Hi Aris:

I believe peak ratings is a good concept. It does discourage sandbagging. It is discouraging to U 2000 players when an expert, temporarily off his game in the prior one or two tournaments, has just slipped below 2000, and decides to play in the U 2000. He may be too strong for the section, and kicking him upstairs would seem to be fair enough.

I think it should be the peak rating in the last 2 years ( in any of the ratings systems ). I like the 2 prior years span, but I suppose it could be lessened to the prior year.

Just my 2 cents worth.

Bob

Hi Bob, but is there an easy way for me to find the peak rating over a certain time period, like 2 years? Excuse my ignorance, but I thought we each had a lifetime peak?

[Bob Armstrong]

Re: Peak Rating Qualification?

Hi Aris:

You are definitely right that the ratings page gives a player, a lifetime peak. But I do not favour using that peak rating. It may well be dated - for example my peak in 2001 was 1911. Over the last 8 years, I have not again gone over 1900. It would seem unfair NOW to saddle me with that rating.

I feel 2 years is a reasonable stretch where someone should still be somewhat within range of a peak rating at that time. It would not be expected that someone would likely deteriorate that fast, under normal cirumstances.

You do raise a good question on practicality of a 2-year peak rating qualification. You would be able to canvas a players' rating page and see his last two years ratings, and find the peak one. This could be done ( with a bit of labour time ) for advanced entries. But I admit, I don't know how an organizer could be looking up this for on-site entries, when he is desperately trying to get Rd. 1 to start on time, and needs to do pairings ASAP ( and would have to guarantee each player was in their proper group ). Having a number of volunteers on-site at registration with internet access to divide up the work would help ( I don't know how realistic that would be ).

I'd have to defer to your experience as an organizer, as to whether there would be any possible answer to the late on-site entries ( your pleas for players to e-mail you and pay later may also help to cut down the number of unanticipated on-site entries ).

Bob

[John Brown]

Re: poll by TD: predefining the sections

I don't really like the idea of players moving up and down between sections to
make pairings. This happened in the Ontario Open this year.
I could never keep track of who was leading the sections and who were floating.

Small tournaments should just be Open events and if you want the first two rounds to be accelerated so there are no mismatched pairings in the early rounds that's ok.

You can award sectional prizes based on the number on entries in a given section. ie if the U2000 is the biggest then give more prizes not necessarily
bigger Prizes than the open but more smaller ones.
So you can give First - 5th for the Tournament and say an U2000,U1800,U1400
with second prizes if money is available. You could always award gift certificates for second prizes from Chapters or similar places that are almost everywhere.

Predefining Sections some times limit your number of players
If everyone who enters knows they have a shot at 1st -5th prizes and maybe a sectional prize then your numbers may go up. It's like having two shots at the same target (a potential prize ) if you do well.

[Aris Marghetis]

Re: poll by TD: predefining the sections

I don't really like the idea of players moving up and down between sections to
make pairings. This happened in the Ontario Open this year.
I could never keep track of who was leading the sections and who were floating.

Small tournaments should just be Open events and if you want the first two rounds to be accelerated so there are no mismatched pairings in the early rounds that's ok.

You can award sectional prizes based on the number on entries in a given section. ie if the U2000 is the biggest then give more prizes not necessarily
bigger Prizes than the open but more smaller ones.
So you can give First - 5th for the Tournament and say an U2000,U1800,U1400
with second prizes if money is available. You could always award gift certificates for second prizes from Chapters or similar places that are almost everywhere.

Predefining Sections some times limit your number of players
If everyone who enters knows they have a shot at 1st -5th prizes and maybe a sectional prize then your numbers may go up. It's like having two shots at the same target (a potential prize ) if you do well.

Yeah, the idea of floating a couple of Amateur players between sections was a scheme to avoid forced byes, as the rest of each section (especially non-Amateurs) had to be paired properly. What I do now is have a dedicated floater per section, which worked last weekend, as everyone played all of their games within their own dedicated section, with no forced byes, because if someone had to sit out a round, it was simply a floater.

Thanks for your feedback though, it was something I wanted to improve for a while.

Back to peak rating ideas, Bob's peak-during-the-last-2-years idea sounds reasonable, but there is no way that it could be calculated onsite for last-minute entries, so I have to let that go. Now, my biggest concern is that if I put people in sections based on their peak rating, that this would increase the number of Amateur players. As an example, my rating is 1945 but peaked at 2048. Note that I believe my actual playing strength is the lower of those two. So, if I were to play in U2000, I would think that I had a decent chance of winning, but if I were to play in U2200, then I might be inclined to Amateur.

So, how about this new idea using peak ratings :

How about if people are put in their current section based on their current rating, but any player can upgrade to the section of their peak rating (I currently have do not have an upgrade option, as some higher-rated players do not like it when the average rating of their opponents is dragged down by upgraded lower-rated players). The other thing is that the upgrading player could not play as an Amateur in the higher section. I have found in the past that if I do not have that clause, then I get some players upgrading who really should not, lose all their games, but most importantly, take away from the tournament experience of their 5 opponents, who each deserve 5 competitive games.

Thanks in advance for any feedback on this peak-rating/upgrading-section proposal.

Yours in chess, Aris.

[Sanjiv Kalra]

Re: poll by TD: predefining the sections

I think people should be able to upgrade to a higher section if their current or peak rating has been within 100 rating points of the higher section. At the same time in all fairness to the players playing in the higher section I agree with
"The other thing is that the upgrading player could not play as an Amateur in the higher section.". I have seen this in Kitchener, where you can play one notch up if you fall within 100 points of it. I am not sure if one had to pay more in Kitchener to do this.

[Egidijus Zeromskis]

Re: poll by TD: predefining the sections

Aris, how was your interview?

Sections should be defined before a tournament and placed in the announcement (with a prize/money distribution).

How many sections:
Open (truly open - all who wants can play; otherwise call it Master's)
Then check your stats and find optimal player numbers for the lower sections and stuck to that for several tournaments. If it does not work, you would hear that ;)

Peak rating:
As I understood, the CFC gave a big boost to the rating several years ago, thus the peak rating might be a number without value. 1 -or 2-year max rating - then you'll need to support the rating system upgrade :D

[Aris Marghetis]

Re: poll by TD: predefining the sections

I think people should be able to upgrade to a higher section if their current or peak rating has been within 100 rating points of the higher section. At the same time in all fairness to the players playing in the higher section I agree with
"The other thing is that the upgrading player could not play as an Amateur in the higher section.". I have seen this in Kitchener, where you can play one notch up if you fall within 100 points of it. I am not sure if one had to pay more in Kitchener to do this.

Hi Sanjiv, thanks for your feedback. Agastya seemed to have some good games this weekend, and I find his tactics improving all the time! I hope he enjoyed his playing.

Going back to your post, there are 3 points in there. We seem to agree on the second point that upgrading players cannot play as Amateur in the higher section, which means I did a decent job of explaining my rationale, and hopefully most players agree with us.

On the first point, I guess we are only 100 rating points apart on this. I guess my rationale is that the player has to have achieved that rating, thus demonstrating that they previously have that upper section experience. Don't forget that I'm not concerned about a 1515 player playing a 1615 player, but rather that when that 1515 player plays up in U2000, he could be facing someone with a 1945 rating (my rating as an example) which is now a spread of over 400 points. So, I still lean towards sticking with the player having already achieved a rating within a section in order to be able to upgrade to it.

Regarding your third point about paying to upgrade, when I had it in the past, it can be a pain when I audit my entry fee totals to determine the prize pool, etc. I find it easier to make the "upgrade payment" simply the requirement to pay as non-Amateur in the higher section. By having a peak rating already in that higher section, they have also in a sense demonstrated that they should be accepted for that section without "penalty".

I hope that my long-winded responses make sense, and look forward to any comments.

Thanks again, and best regards, Aris.