WCH Stage 1 groups new players

  

Back to forum


Garvin Gray    (2010-09-30)
WCH Stage 1 groups (new players)

I wish to propose that the stage one groups starting from the next WCH tournament be paired differently.

Currently, from my understanding, all the players who have not qualified for the high rated round robins or elimination matches are paired into separate round robins of about seven players in each group.

Seed number 1 is in group 1 and so forth until all groups have been allocated.

The idea of this being to try and ensure that each of the groups is of equal strength.

Where I think this falls down is the issue of players with provisional ratings ie players with new ratings of 1800, 1500.

I have had the experience of having one or more of these 1800's in my group and after the group is finished, it is clear that the 1800 player has achieved a rating of 2100 plus, meaning that my group had three players with playing ability over 2100, meaning my qual group was unfairly disadvantaged.

How I would like the stage one qualifying to work from now is:

1) All players with recognised ratings are paired as per normal.
2) All the provisional rated players are put into groups by themselves.

Then normal qualification rules apply for getting to stage 2.

It is highly unlikely that a low rated player will qualify from the provisional rated groups as someone from each of those groups will be about 2100 or so by the end of the first qualifying stage.

I do ask for this to be endorsed for the next WCH.


Daniel Parmet    (2010-09-30 21:56:55)
WCH Stage 1 groups (new players)

I agree immensely!

Further i've noticed that the color pairings tends to be very very bad for seed 2 & 3 but great for seed 1 & 4.

As seed 2 I get black against 1&4 as seed 3 I get black against 1 4 but as seed 4 I get white against 1&3. Yet these color disadvantages are not at all included as a tiebreak in any way.

The result is that in the last Wch as seed2 I never got a useful white having two tough blacks with seed #1 & #4 but in the Wch before as seed #4 I have no tough blacks because I had white against seed #1 &#3. I'm not quite sure the solution other than to somehow tie the colors into tiebreak before ratings.


Thibault de Vassal    (2010-10-01 01:42:40)
WCH Stage 1 groups (new players)

Hi Daniel, maybe it would be more fair to reverse the colors (seed 1 plays black against seed 2 and so on)... I'm not sure if it is better but we can think about it. The color is not a good tiebreak IMO, does anyone else have an opinion on this?

Hi Garvin : "2) All the provisional rated players are put into groups by themselves." , you mean they play together in special groups? We can think about it as well but one goal of the championship was to help those players to find their place quicker in the rating list before the next cycles. I'm not sure if a 2300 player provisionnaly rated 1800 is an advantage for anyone else in the group more than seed 1.


Garvin Gray    (2010-10-01 13:21:12)
WCH Stage 1 groups (new players)

Colour allocation- There are two answers to this, the most pure solution is to have double round robin groups. If this means six player groups and more groups in total, so be it, but at least it would solve the issue.

Berger Pairing Tables- It seems that the round robin groups are not paired using the Berger Pairing Tables for single round robins, why not?

If they are not, then they should be, solves all these issues of colours. This should be the case for all round robins on this site.


Garvin Gray    (2010-10-01 13:30:27)
WCH Stage 1 groups (new players)

GG- "2) All the provisional rated players are put into groups by themselves." ,

TB- You mean they play together in special groups?
We can think about it as well but one goal of the championship was to help those players to find their place quicker in the rating list before the next cycles. I'm not sure if a 2300 player provisionnaly rated 1800 is an advantage for anyone else in the group more than seed 1.

GG- It is not an advantage to have an 1800 in your group if they play to a standard of 2100. It is a severe disadvantage.

It means there is one more person in some groups that plays to a rating way above their provisional rating.

I am very concerned that you seem to be putting the needs of increasing those players ratings in the WCH above the integrity of the competition as a whole. It means you are unfairly affecting other players chances of qualifying, just for the sake of allowing new members the chance to gain a few extra rating points.

The new members still have a lot of chances to increase their rating through playing in normal tournaments, which is where the longer term members had to get their ratings from.

I am saying that those with provisional ratings should be seeded into groups by themselves in stage one.

Whoever wins these groups will clearly be about 2100/2200 playing strength and so will not be crushed in stage two anymore than those with long term 2100 ratings.

A secondary option is to seed some of these players using their advanced rating (if they have one), so at least then there does not end up being three or four 2100's trying to qualify from the same group, while having other groups with only one or two 2100's.


Daniel Parmet    (2010-10-01 17:36:43)
WCH Stage 1 groups (new players)

I agree with everything Garvin said.

I would also like to note that I agree DRR would be great to solve the color issue but understand if thats not possible.


Thibault de Vassal    (2010-10-01 17:39:15)
WCH Stage 1 groups (new players)

I only meant that it is an advantage to play a 2100 player actually rated 1800 rather than 2100 in his group... Gives more chances to qualify but of course, a few Elo points will probably be lost in the process. There is a matter of luck there of course, just like when a player loses on time or forfeits his games in a group, that's why I try to launch almost 2 new cycles per year, it is not possible to avoid the lucky factor in a championship but it is possible to give more chances to everyone...

Anyway I don't think it is possible to build groups of perfectly equal strength (btw many players rated 2000-2100 after 50+ games are actually worth 2400 and more, but lost many games on time) while we need clear rules (if those players are in special groups, do they qualify the same way?).

Another option would be to prevent players who have not finished a certain number of games to enter the WCH waiting list... :/ I don't think it would be appreciated.


Thibault de Vassal    (2010-10-01 17:51:59)
On colour allocation

On colour allocations, there was numerous discussions on this topic during the 1st year of the server. Well, I cannot remember exactly all my arguments, but briefly 1) Double round-robin is too much effort for the players while it does not eliminate totally the chancy factor. 2) On Berger, the whole FICGS WCH idea is to give more importance to the non-WCH tournaments, the very best player must be champion IMO, not only the winner of a few tournaments, that's why ratings are so important in the tie breaks (and that's why my first idea was to give White to the top seed in round robin groups)!

Less games for everyone per cycle + More cycles = More chances to find the real champion (and more fun :)) !


Scott Nichols    (2010-10-01 19:29:31)
WCH Stage 1 groups (new players)

In my last WCH bracket I was rated 514 points above the avg. of my opponents.

I would like to suggest--All players over 2000 with at least 50 ficgs games played should be "seeded" into round 2. This would give players an incentive to play more and to try to reach the 2000 plateau.

Players below 2000 and less than 50 games would all have to play stage 1 to "qualify" for stage 2.


Thibault de Vassal    (2010-10-02 00:03:55)
WCH Stage 1 groups (new players)

The number of players in each round must be coherent (round 3 : 5 to 9, round 2 : 5 to 9 x 5 to 9, round 1 : 5 to 9 x 5 to 9 x 5 to 9) ... if things are easier for players rated 2000-2300 during the first round, it will be harder for them later (e.g. 1 player qualified in a group of 11 players!).

But it is true by experience that the rating limit for the M group may be below 2300 (by the way it was for the last championship)

I'm against such a rule "All players over 2000 with at least 50 ficgs games played should be seeded into round 2", a bit complicated IMO, and players rated 2250-2400 will have the same kind of complaining then :)


Daniel Parmet    (2010-10-02 00:06:39)
WCH Stage 1 groups (new players)

In response, I don't mind a stage 2 being 11 players if 2000+ was seeded into it.


Scott Nichols    (2010-10-02 13:14:17)
WCH Stage 1 groups (new players)

Can you tell me what the standard is for being seeded into round 2? Then I can just wait until I reach that point (if ever) to join WCH, thank you.


Daniel Parmet    (2010-10-02 18:14:34)
WCH Stage 1 groups (new players)

This is what the rules currently say "Players with a rating superior or equal to 2300 will play 1st stage in high rated groups if possible. Winners of these groups will be directly qualified for stage 3, others will play stage 2."

Though by that, it means that all players have to play stage1 regardless?


Scott Nichols    (2010-10-02 21:45:32)
WCH Stage 1 groups (new players)

I'm sorry Daniel, but it still doesn't seem to be any clearer...


Daniel Parmet    (2010-10-02 21:51:28)
WCH Stage 1 groups (new players)

Yea its not clear to me either. I thought that it meant 2300s get seeded into stage2. No one gets seeded into stage 1. Winners of stage1 have to play stage2 to get into stage3. However this is not at all what it says.

It says EVERYONE plays stage1. If someone 2299+ wins stage1 they can skip stage 2. If someone 2299 or lower wins stage1 they go on to stage 2.


Daniel Parmet    (2010-10-02 21:52:10)
WCH Stage 1 groups (new players)

Err should say "no one gets seeded into stage 2"


Thibault de Vassal    (2010-10-02 23:26:17)
WCH Stage 1 groups (new players)

Hard to answer it. It only means what it means (but if you find a better way to say it, I can update the WCH rules), it can be concluded from the whole text (not from this excerpt only) that everyone plays round 1, indeed.


Daniel Parmet    (2010-10-04 01:02:30)
WCH Stage 1 groups (new players)

I would say a better way of wording it is that Everyone plays stage1. If someone 2299+ wins stage 1 they skip stage2. If someone 2299 or lower wins stage 1 they go on to a stage 2 group.


Thibault de Vassal    (2010-10-04 19:20:13)
WCH Stage 1 groups (new players)

Yes, but it doesn't mention the M groups, this way the number of players in the final round-robin tournament would be quite hard to predict.


Garvin Gray    (2010-10-05 10:54:29)
WCH Stage 1 groups (new players)

Thibault must be impressed. The whole topic has moved away from the orginal discussion point, so that issue has dropped off the radar, meaning changes are not likely while it is not being discussed.


Thibault de Vassal    (2010-10-05 13:09:20)
WCH Stage 1 groups (new players)

Hi Garvin, to continue on your original point, I don't think that a player provisionally rated 1800 in a WCH group is a real problem: It is not a significant advantage for anyone in the tournament (the chancy factor always exists in 7 players tournaments anyway), there is a rule [for ~2 years now] that prevents high rated players to lose many points in case of a loss or draw against such a 1800 player who is actually worth 2300 or more (there are other occasions to get free points btw e.g. general forfeits), and WCH groups help these players to find their real rating quicker. As I said the number of cycles is the point, giving more chances to everyone.

But it is true that the colors (top seed playing White against seed 2) may be reversed, so far the idea was "the rating does count to give more chances to the best player to become champion" but maybe the advantage is too big. It still needs to be discussed though.


Garvin Gray    (2010-10-05 13:36:15)
WCH Stage 1 groups (new players)

Thibault,

You have missed my original point. I am saying that having a 1800 player in your group can be a DISADVANTAGE.

In none of my postings on this topic have I mentioned anything about ratings, except to express a lot of concern that you seem more concern about using the WCH tournament to improve ratings than to try and qualify the best player from each group and to have each of the groups of as close to equal standard as possible.


Daniel Parmet    (2010-10-05 13:56:29)
WCH Stage 1 groups (new players)

I must say I agree with Garvin on all points though I have nothing more constructive to add that he hasn't said.


Thibault de Vassal    (2010-10-05 14:21:22)
WCH Stage 1 groups (new players)

If it is a disadvantage, it should be consequently an advantage for someone else, I meant this way... So your point is that it is a disadvantage "in the tournament", right? I do not agree with this, if the best player was actually this 1800 player, he should be able to play the championship anyway (and you have the advantage of ratings there for tiebreaks)... If players with a provisional rating play together in special wch groups, the winners (probably still under-rated) will play stage 2 and we'll have the same problem then IMO.

Do other players have an opinion or similar arguments on this point?


Garvin Gray    (2010-10-05 16:31:02)
WCH Stage 1 groups (new players)

Thibault, I am more than happy to let others give their opinions, but I really do not think you understand at all what point I am trying to make.

This is now three times that you have misunderstood what I am trying to say.

Geez I wish we could quote better in these forums. It would make discussing points much easier.

Thib: So your point is that it is a disadvantage "in the tournament", right? I do not agree with this, if the best player was actually this 1800 player, he should be able to play the championship anyway (and you have the advantage of ratings there for tiebreaks)...

GG- I am not arguing at all that the 1800's should not be able to play in the championship. Please stop mis-quoting me. I have also stated this previously. I am stating that they should be in groups in stage one all by themselves.

The disadvantage is with how the groups are paired and I finding it very difficult to not get completely pissed off with having to explain items many times for you to understand what I am trying to say.

You keep failing to respond directly to my points and I keep having to point out how you have mis-quoted my points, which does not help in the debate at all.

The groups are currently paired in the first stage with the highest rated player in Group 1, second highest rated player in Group 2, third highest rated player in Group 3 and so forth for eleven groups (in this example there are eleven groups). Then the 12th highest rated player is placed in Group 11, the thirteen highest rated player in Group 10 and back we go to the 22nd highest rated player in Group 1. The pattern keeps repeating back and forth until all players in stage one have been allocated to a Group.

Now with the 1800's being seeded in these groups with their 1800 rating is that they end up being about the 4th or 5th seed in some groups, but are not allocated to each group.

Now when some of these 1800 players start performing at a rating of 2100, it means in some groups that the top seeds have received three players of similar playing level and some other groups have not. This makes some of the groups disproportionately unfair.

If these 1800 players were somehow seeded accurately according to their playing standard, meaning they entered stage one in their proper seeding position, it would push all the rest of the players down one spot and so the Group allocations would be fairer.

Another option could also be to make it a rule that players must have a proper rating ie not provisional, before being able to play in the championship. I have tried to avoid suggesting this with my proposal to have them play in a group all by themselves.


Thibault de Vassal    (2010-10-05 17:09:25)
WCH Stage 1 groups (new players)

Aaaaaah... makes me mad :/ Okay, one more time:

GG- I am not arguing at all that the 1800's should not be able to play in the championship. Please stop mis-quoting me. I have also stated this previously. I am stating that they should be in groups in stage one all by themselves.

Thib- I just said "if players with a provisional rating play together in special wch groups", that's the way I (mis-)understood you, so I did not say that you argued that they shouldn't play wch... The thing that is not clear to me is "I am stating that they should be in groups in stage one all by themselves.", so does it mean:

1) Provisionnaly rated players should play in special groups
2) There should be the same number of provisionnaly rated players per group
3) Tie break shouldn't be TER for provisionnaly rated players
4) Provis. rated players should earn their WCH entry, eg. # ended games >50
5) Something else

About cases 1), 3) & 4) I answered it in the discussion that I just re-read entirely.

Sorry for mis-understanding you & thanks for your patience... If I still completely miss it, maybe better would be that another player try to explain me what you mean by "they should be in groups in stage one all by themselves".


Philip Roe    (2010-10-05 17:20:42)
WCH Stage 1 groups (new players)

I think that the management of the WCH should not be too heavily weighted toward ensuring that "the best player" wins. On behalf of the underdogs, I would like us to have at least a sporting chance. If the cards are too much stacked against us the idea of an "open" tournament is lost, and we won't enter.

I looked at the statistics for cycle 000007. The top seed won outright 7 times, and tied for first on 7 other occasions. The second seed won outright twice, and tied first 7 times. The third seed won 5 times and tied twice. The fourth seed won once and tied twice. Out of all the winners, only the the two fourth seeds who tied had provisional 1800 ratings.

Are these numbers really a cause for concern?


Thibault de Vassal    (2010-10-05 17:58:43)
WCH Stage 1 groups (new players)

Hi Philip, thanks for taking some time to give us these numbers, really appreciated!

Well, I did not hide that I thought about this championship this way, simply because I wanted it also to look like the old classical chess championship. The point that is discussed here is a tiny detail only compared to the whole idea... Of course the 8 players of the knockout cycle have much better chances to reach the final, and the current champion is by far the favourite. Why to play a WCH that would be a boring copy of IECG & ICCF WCH?

I think that everyone has a real chance though, maybe hardly on 1 cycle but by playing 2 or 3! Have a look at Edward Kotlyanskiy's tournaments, he started WCH 3 with a rating of 2132, seed #2 in a RR group... he is champion!

http://www.ficgs.com/tournament_FICGS__CHESS__WCH_STAGE_1_GROUP_10__000003.html


Daniel Parmet    (2010-10-05 19:12:30)
WCH Stage 1 groups (new players)

My point is that seed one has a HUGE advantage because of the fact 1) Tie breakers AND 2) colors are always in their favor. So it makes it very very hard for another player to overturn this. Black against the highest rated usually best you can ask for is a draw resulting in the fact you now have to win the rest of your games in order to not Tie the highest rated (or pray for an upset). Cause if you tie... they just win on tiebreaks.


Thibault de Vassal    (2010-10-05 19:36:41)
WCH Stage 1 groups (new players)

Well, according to the numbers given by Philip, the number of wins by a player who is not seed 1 is over the number of wins by seed 1, but I agree that the advantage is big, that's why we could reverse the colors (I'll open a new topic to discuss it if we continue this way). But I still think that 7 players are open enough, everything can happen.


Daniel Parmet    (2010-10-05 20:41:54)
WCH Stage 1 groups (new players)

No according to the numbers given by Philip my point is huge.

7 wins, 7 ties meaning 14 went on.
2 wins, 7 ties meaning 2 went on.
5 wins, 2 ties meaning 5 went on.
1 win, 2 ties meaning 1 went on.

Notice that because of color and tiebreaks the giant separation in place 1 2 and 3. It actually went back on spot 3 because they had white most likely. Clearly spot 2 is the worst position to hold in the tournament. Everything is against you. Color and tiebreaks.

Anyways, I agree it needs to be discussed. If others disagree with me then thats that I guess.


Thibault de Vassal    (2010-10-05 21:22:04)
WCH Stage 1 groups (new players)

Okay, the numbers must be taken carefully, 2 wins 7 ties for seed 2 do not necessarily mean 2 went on... And I just counted for WCH 6:

Seed 1 - 2 wins 4 ties 6 qualified
Seed 2 - 2 wins 1 ties 3 qualified
Seed 3 - 5 wins 3 ties 5 qualified
Seed 4 - 1 wins 0 ties 1 qualified
Seed 5 - 0 wins 1 ties 1 qualified
Seed 6 - 0 wins 1 ties 0 qualified
Seed 7 - 1 wins 0 ties 1 qualified

And for WCH 5:

Seed 1 - 2 wins 2 ties 4 qualified
Seed 2 - 3 wins 2 ties 3 qualified
Seed 3 - 2 wins 3 ties 4 qualified
Seed 4 - 0 wins 0 ties 0 qualified
Seed 5 - 2 wins 2 ties 2 qualified
Seed 6 - 2 wins 0 ties 2 qualified

Definitely everything can happen in these groups, and I did not count the players invited to join stage 2, that do not favour top seed and that actually favour seed 2 over seed 3... Anyway as I said, I'm ok to discuss the idea to reverse colors in a new topic.


Ruben Comes    (2010-10-10 07:00:31)
WCH Stage 1 groups (new players)

Guys, beyond the technical, I think that is a world champion, be able to pick any group players have 1800 or 2100. This is my humble opinion.


Lalit Kapoor    (2010-10-20 08:18:46)
WCH Stage 1 groups (new players)

The same happens to me. I started as 1805 (provisional) in WCH 8. Now after 5 games finished I scored 4.5/5. Against 2nd seed I got a draw. And against seed 1 I will get at least a draw with slight advantage as Black in endgame.

Garvin Gray question is right. The 1800 (provisional) player gets an advantage for being an underdog.

But at last there would be only one champion and he have to qualify all stages.


Garvin Gray    (2010-10-20 08:42:43)
WCH Stage 1 groups (new players)

Wow, finally a new post. :)

I am not sure if you have received an advantage by being in a group with other fully rated players.

My premise is that you should have been a group, or groups, with other provisionally rated players, so that those who have worked to get an accurate rating get to try and qualify in roughly even strength groups (in comparison to all the other qualifying groups).

As someone who has shown themselves to be of decent standard, you would have still probably qualified through that group of provisionals and most likely would not be out of your depth in stage 2.

Which again is the whole premise of my position.