|
|
|
Topic: Semifinal assignments in Auto Open events
| |
|
|
Author
| Message |
|---|
|
|
I am on of 5 players who made the semifinals of open 2692.
The 3 players above 2250 are in one group. the 2 players below 1800 are in another group. This does not make any sense. In soccer doesnt the top seed play #16. The the winner of 1-16 plays the winner of 8-9. 1 and 2 dont meet until the finals.
When the are 2 groups in the semifinal I think the should be filled as follows.
Group 1
1, 4, 5, 8
Group 2
2, 3, 6, 7
does anyone know what the standard practice in chess and other sports is when you want 4 player groups iun the first round and up to 4 players per group in round 2? If we can agree on what is best I will post that in feature request.
|
|
|
The problem on many chess-sites is that the line-up for the group stages and semi-finals is done by the computer. Also at Queen Alice.
Last month I reaced twice the semi-final of RTC tournaments, 10 min/game, in which of both occations I managed to draw, against opponents ratewd more then 200p. as me.
Because of their rating, in both occations, they went to the final, which to my point of view is very unfair.
More logical would be that the lower rated competetar goes trough, or the one who plays the black pieces.
|
|
|
MrDJRVD it is odd that you would win but not advanced. Atleast we dont have that problem.
The semifinal groups are etermined by computer. I can write a program that would put the top seed in group 1 and the second seed in group 2.
The first criteria for seed 1 should be performance in round 1. to be logically consistent round 1 groups should be seeded by rating. Does any one know the proper seeding for round 1? my guess would be
group A 1,10,11,20
group B 2,9,12,19
Group C 3,8,13,18
Group D 4,7,14,17
Group E 5,6,15,16
|
|
|
In chess postal calculate the average rating for each group.
Considering: 1 (2280), 2 (2386), 3 (2335), 4 (1785) and 5 (1585) would:
Group A = 2 and 4 (average rating: 2085.5) and Group B = 3, 1 and 5 (average rating: 2066.7).
|
|
|
Dogeval
Your idea is a great improvement over what we hapresently have. It is more difficult to program. Unlike an OTB tournament raatngs change during the tournament? I suppose program would use present ratings at the time of forming the groups. I think that ratings should only influence assignment of initial pairring number. My groups assume that players were seeded in round 1. Your idea would be better for tournaments already started. For new tournaments if the original seed 2 score 12-0 in round 1 and the original seed 1 score 11-1 then the original seed 2 would become the top seed for round 2.
|
|
|
|
|
