One tournament for each astrological signs of the Zodiac. (double elimination)
Be carreful about the time settings. Use the conditional move feature to speed up the game. (Canadian Byo-Yomi: Clock starts with 2 days main time, followed by 1 week per 28 stones.)
No pausing and no vacations; anyone who has vacation mode switched on, or pauses a game, will be disqualified.
Also, please do not use AI.
Title holder :
Lion :
Virgo :
Libra :
“Players: 4-100; Tournament type: Double elimination.”
With 16 players, a double-elimination tournament would have 10 or 11 rounds. A double-elimination tournament with 100 players would need 16 or 17 rounds, if I count correctly.
One round should last about two months with these time settings (250 moves per game, divided by 28 moves per week).
I fix the maximum number of player to 16.
With 16 players, I get 7 or 8 rounds for a double-elimination tournament (rather than 10 or 11 rounds).
after # of # of players in # of players in
rounds finished winners bracket losers bracket
0 16 0
1 8 8
2 4 8
3 2 6
4 1 4
5 1 2
6 1 1
7 0 or 1 0 or 2
8 0 or 1 0 or 1
​
(Of course, if one wants it to take more rounds, one can get that
by reducing the number of games in some or all of the rounds.)
@hoctaph You and I might not be using the same definition of “rounds” when counting them.
Perhaps more specifically, I should have said:
With 16 players in a double-elimination tournament, if a player loses their first game then wins all subsequent games, then this player will play a total of 11 games in the tournament:
I don’t know exactly how you divide all the games played during the tournaments in “rounds”; there might be more than one way; but in any case, the player described above cannot start their next game before they have finished their current game.
EDIT: Actually only one game in the final, for the reason pointed out by @teapoweredrobot said below. So, 10 consecutive games for a 16-player double-elimination tournament, and 16 consecutive games for a 100-player double-elimination tournament.
But not on OGS…
Double elimination is in TATTERS
You Should Have to Lose Twice to Lose Double Elimination
Etc…
One could have such a player play a total of 11 games in the tournament, via
1 or more of the rounds have more than 1 of the [players who have
not yet been eliminated] not play against another of those players
or
1 or more of [the games between those players] before the finals are
[someone in the winner’s bracket] against [someone in the loser’s bracket]
, ​ but if one doesn’t slow things down, then no player would play more than 8 games,
even when using actual double-elimination rather than this site’s version:
​
​
16 in WB , 0 in LB
each player plays 1 game
8 in WB , 8 in LB
each player plays 1 game
4 in WB , 8 in LB
each of the above 12 plays 1 game
2 in WB , 6 in LB
each of the above 8 plays 1 game
1 in WB , 4 in LB
the 4 players in LB each play 1 game
1 in WB , 2 in LB
the 2 players in LB each play 1 game
1 in WB , 1 in LB
1-or-2-game finals(I believe this one is back on-topic, because it could result in
@TimSeoknawibo changing how the tournament will be run.)
With 8 rounds, actual double-elimination (i.e., not what this site does when
that option is chosen) can in fact handle 20 players, rather than just 16.
Furthermore, still with just 8 rounds, one can handle [20 players who need to lose twice to
be eliminated] plus [4 players who get eliminated by a single loss] (analogous to McMahon).
For 65 players, actual double-elimination can need 11 rounds, and 11 rounds lets it handle 128 players.
(though here, adding any who can be eliminated with 1 loss
requires reducing the number who need 2 loses to be eliminated)
def yield_nextsibs_(winners,losers):
if (winners+losers)%2 == 0:
byesibs = ((0,0),)
elif losers%2 == 1:
byesibs = (((0,1),) if winners == 0 else ((0,1),(1,0)))
else:
byesibs = (((1,0),) if losers == 0 else ((1,0),(0,1)))
for byes in byesibs:
winnerswhoplay,loserswhoplay = winners-byes[0],losers-byes[1]
cross_sibs = (((1,0),(0,2)) if loserswhoplay%2 == 1 else ((0,0),))
yield tuple(((winnerswhoplay//2)+byes[0]+crossresult[0],(winnerswhoplay//2)+(loserswhoplay//2)+byes[1]+crossresult[1]) for crossresult in cross_sibs)
roundsneeded = [[-1 for losers in range(140)] for winners in range(140)]
roundsneeded[0][1],roundsneeded[1][0] = 0,0
for total in range(2,140):
for winners in range(total+1):
losers = total-winners
roundsneeded[winners][losers] = 1+min(max(roundsneeded[W][L] for (W,L) in byechoice) for byechoice in yield_nextsibs_(winners,losers))
print((roundsneeded[12][4],roundsneeded[12][5],roundsneeded[13][0]))
print((roundsneeded[20][4],roundsneeded[20][5],roundsneeded[21][0]))
print((roundsneeded[64][64],roundsneeded[64][65],roundsneeded[65][0]))
print((roundsneeded[128][0],roundsneeded[128][1],roundsneeded[129][0]))
Letting OGS handle the tournament is much smoother than running the tournament by hand, if the number of games is large. In particular I think the games start automatically, as opposed to the tournament organiser having to harass the late players to remind them to start their games.
Also, I should note that the way you described the rounds for 16 players is pretty different from what I’ve seen done elsewhere.
For instance: https://upload.wikimedia.org/wikipedia/commons/b/b0/NSB-doubleelim-draw-2004.svg
Well, if the final is the only issue with what OGS’s system
does when double-elimination is chosen, then one can do
.
​
​
However, I guess there’s a risk that the system will give pairings which
increase the number of rounds earlier in the tournament (I imagine that the
bracket @ArsenLapin1 linked to is an example of such pairings), despite
sacrificing double-elimination to sometimes save 1 round at the very end.
Is there a way to see the pairings or the brackets for the tournament?
The page at Zodiac : Libra Tournament doesn’t appear to show anything in that regard, not even links to the games currently being played.
The icons on the left link to the games. It’s first in that list against second and so on. Not ideal even with just 7 players. With more players you have no clue if a specific player is paired against the person above or below. So yes, this definitely could be improved. But personally I think getting back the ordering by points in McMhaon and swiss tournaments is even more important. (You didn’t ask for this, but I’ll take any chance to mention it anyway 
)
