Thad: The BKR value at the moment of accepting the draw is always used.

Just thinking here. If the lower seated player is to advance in the case of a draw, what happens if Al & Bob are playing a tournament match and Al is higher at the start, but then Al loses a regular game to Carl and his rating goes down to below what Bob's is. Al's was higher at the start of the match, but Bob's was higher at the end. So who advances?

Sorry, I don't have a good solution to this, just thought it should be brought up.

Single Elimination Backgammon #2

6 players still needed - Tourney starts 1 day 19 hours from now.

*added* but I will start it earlier when it fills up.

Very fast 4 players elimination tournament here

Coudy for winning the tournament Backgammon & Hyper 2100+ First 8 (Hyper Backgammon)

furbster: If it would be in a final, I could even lose a game on purpose, just to put my BKR lower, so I could win that tournament... :(

It seems to be that this is not a solution... Imagine a prize tourney being dependent of player's BKR...

furbster: Do not let them play against eachother in the first round. The player who beats the strongest player in the first round wins.

rabbitoid: but what happens if the ratings are so close or even the same eg, 1501 : 1502

rabbitoid: Yes, that's a good idea.

I played last month such a tournament and indeed if a low rated player draws a higher rated player, then the latter on losts. In the first rounds the rating differences were big, but the semi-final was close. (I was knocked out in the quarter final)

Thad: This isn't a problem that only occurs at single elimination unless you play 2 game matches in a normal tournament. By the way, if a game is too unfair (which Chess is certainly not) I wouldn't play it ;-)

rabbitoid: That still leaves a problem in games line Pente where P1 enjoys a strong advantage. For average players, this doesn't really matter, but just as in Chess where strong play on both sides results in a draw, strong play by both sides will produce a win for P1.

Fencer: instead of a "no draw" rule, which impossible to impose in chess, why not fix that in case of a draw the lower rated player stays?

J.T. single elim.I

Luke Skywalker:
I would expect only one, or possibly two if a single-elimination tournament can be defined with two games between each pair of opponents.

But I've observed that the number of slots they tell you to need often has little or nothing to do with the number of slots you actually need. Caveat emptor.

i believe it will be unwise to make single elimination tourney for games boards(chess or checkers) because the possibilities to have draws are high especially on the final rounds unless is used the bkr :)

Luke Skywalker: Game slots are not used for single elimination tournaments. Because you'll never play more than 1 game at the same moment.

Lambik: I might change it.

how many game slots will pawns and knights need for this kind of tournament?

Fencer: With single elimination, it is not possible to "draw" a game. But how about games that could not be won? For example a chess game with not enough material (K+N+N against K for example) or what about the fifty moves rule? Some time ago I had a game with three times the same position, but no automatic draw. Is this doable?

The Sunshine Club has lots of new tournaments, and some Ponds, for anyone interested!


http://brainking.com/en/ShowFellowship?fid=41

Fencer: oh cool!

I didn't realize that they will start as filled! That is awesome!!!!

Luke Skywalker: Yes, this is a known issue. A side effect of new features :-)

I have 2 tourneys set up. Lets try them out!


Single Elimination Backgammon
(set up for 4 players each....)
# Backgammon
# Nackgammon
# Crowded Backgammon
# Hyper Backgammon

Single Elimination Backgammon #2
16 players needed

Luke Skywalker: very good point, in fact i've just been looking for it, adn i decided it'd been deleted since my lunchbreak.

this tourney
http://brainking.com/en/Tournaments?trg=13412&trnst=0
has already started, but not enough people have singed up for some of the games. When the tourn was still in the "signed" state, it stated a deadline of 29 days until those would be deleted.
2 problems:
- the deadline is not stated now
- the tourn is not listed in the "signed" category anymore, so people probably won't find it and won't know that they still can sign up for it.

http://brainking.com/en/Tournaments?trg=13452&trnst=0

Who wants a free membership?

Fencer: aha! you clever devil! (and excellent coder, i might add! ;)

plaintiger: It is currently disabled because I need to test it first. But I'll enable it soon (by uncommenting one line in the source code ).

question: how does one *start* a tournament with an entry fee? i don't see any options for it in the tournament setup pages. thanks...

Fencer: Oh, then I misunderstood this ;) By the way, it's nice to see the tournaments marked after joining.

Retep: It was just an example. Nobody says the prize must be always only 25%. The tournament I've just created has 40% prize for the winner.

Doesn't matter.

I agree with arpa, that's very "unpleasant"! This:

# The tournament will not be started until the first prize reaches at least a double of the entry fee. Players are informed about this status on the tournament page. #

means a prize tournament can't be started with less than 8 players.

Fencer: 1) tomorrow? i don't know this! wonderful!

2) oops ... i change that word in "unpleasant" ok?

arpa: Why don't you wait for tomorrow? The S-B is already fixed, it just needs to be uploaded.
Btw, I don't like the word "stealing".

see Help page

8.6.2) Prize tournaments with an entry fee

instead of .... The tournament will not be started until the first prize reaches at least a double of the entry fee.

8 players it's the same.

same page ....The tournament is defined with these prizes: 25% to the winner, 10% to the second place and 5% to the third place....

I think 60% "loss" is an stealing ... 20% are reasonable (50% 20% 10%).

NECESSARY REWIEW S-B CALCULATION (today are incorrect)!!! see(for example):

http://senseis.xmp.net/?SonnebornBerger

arpa: Huh?

at Dark Battleboats in the tourney Scarlets Springtime Follies".

Sider line 4 first 8 to sign, two games each, under 2000 rating.


http://brainking.com/en/Tournaments?trg=13388&tri=83887&trnst=0

Walter Montego: I am not working on it. But I plan to do so.

Fencer: This will be a very good improvement. I'm glad you're working on it. Good luck.

Congrats to CameronsDad for winning Speed Gammon #2's Backgammon Race!

Walter Montego: Simply said, when it's clear than one (and only one) player would have more points than anybody else, regardless of results of the unfinished games of the same section, he can be declared as the winner. In all other cases when S-B points could affect the final order, it's more complicated and it's safer to wait until everything is completed.

Fencer: I was thinking along the lines of having the program check the tournament after one person has finished all of the games. As you say, if that person has won all their games, they're the winner of the section. For all the other situations you could have a chart for each possibilty and have it check the chart. It might be easier to write a program to create the chart and then just use the chart as a table look up kind of thing. This seems like a lot of work to me and you'd need a different chart for each size section. Still, once there's a chart covering every way a section can be finished, it'd check it fast and wouldn't require any further calculations.

I'm thinking a brute force method might be the way to go even though it requires a calculation each time. There's not that many different ways a tournament can finish after one person has finished all of his games. You could just have it check to see if that person wins no matter what happens to the other people in the tournament. The moment it comes back with someone else the search is over until the next completed game and then check each person that has finished every game in the section.

If all that is too much trouble, it would still be a good thing just to have it cover some very special cases. The case with one person being done and having won every game. The cases where one person is done and only lost one game and everyone else has at least two losses or has lost to this person. Just adding this would probably cut down on a lot of idle tournaments, let alone if you covered every possibility.

I'm thinking the "two games colors switched with drawn games counting as a half point each" kind of tournament would take a different set up, or would it be the same thing?

Walter Montego: It's easy to cover simple situations where one players wins all games in his sections - then nobody else can be the section winner. In other cases it can be a little more complicated, as you say, many "if this, than that". It's not the biggest priority to implement it now but I will think about it.

Fencer: You'll have to have it either end the tournament or if there's more than one section, advance the winners to the next round if the other sections are waiting. The players whose games don't affect the winners of a section can continue to play their game, but everyone still in the tournament can get playing again. Or the tournament winner is now known and those slower players can finish their game. I'm sure Pawn and Knight members will like this improvement too. They'll be free to play in another tournament while the slowpokes can play their game in peace.

It sounds like a lot of "if this, then that" kind of programming. You'll have to take into account a lot of different scenarios to insure getting every situation covered. I'm sure there's a finite amount of ways a section can be completed. Or would a brute force method work and you just plug every win or loss possible and check the value for each after each game is completed? If it comes out the same for all situations it'd be time to advance or announce winners, right?

gringo: Because the current system is working like that. The new one might have a more sophisticated algorithm.