From: "Frank Hamrick"
Subject: RE: [BCML] Ways to handle unbalanced scenarios in a tournament
Here's my proposal for a six-week, 8-player, balanced Battle Cry tournament:
The Gettysburg Campaign.
The first week -
Scenario - McPherson's Ridge
Game one - Players A1, A2, A3, A4 are CSA; Players B1, B2, B3, B4 are US.
A1 vs. B1; A2 vs. B2; A3 vs. B3; A4 vs. B4.
Game two - same scenario, but players move "down the table" as you
mentioned:
A1vs. B2; A2 vs. B3; A3 vs. B4; A4 vs. B1, etc.
Games three & four - players continue moving until all A players have played
all B players.
Depending on the game length - this ends the first "tournament night" (but
doesn't end the tournament).
The second week -
Scenario - McPherson's Ridge
Game one - A players are US; B players are CSA
Once more each person plays 4 games as in the previous week
Games two, three, four - players repeat week one schedule, but are each
playing the opposite side
The third week -
Scenario - Devil's Den
Repeat the same procedure as in the first week.
The fourth week -
Scenario - Devil's Den
Repeat the same procedure as in the second week.
The fifth week -
Scenario - Pickett's Charge
Repeat the same procedure as in the first week
The sixth week -
Scenario - Pickett's Charge
Repeat the same procedure as in the second week
I realize this is a long time - and the problem is that some of our guys
may have a problem meeting each week, but they should promise to "make up"
the game if they miss the normal game night, before the next week's
games, thus keeping everyone on schedule.
Total flags won is the overall winner.
You don't have to worry about scenario balance, as everyone plays everyone
on the opposite sides.
Finally - consider taking the A winner and letting him play the B winner
in a championship game or games.
Frank Hamrick
-----Original Message-----
From: Russ Williams [mailto:russw@austin.rr.com]
Sent: Thursday, September 28, 2000 10:30 AM
To: battlecry2000@egroups.com
Subject: [BCML] Ways to handle unbalanced scenarios in a tournament
Hi folks,
I already posted this to rec.games.board, but I thought it worthwhile to
post it to this list as well in case anyone missed it, and some of you might
have some Battle Cry-specific feedback/ideas... I've been talking about this
stuff with Brady, who's running a tournament this weekend!
thanks,
russ
-----
My friend JP & I were talking about ways to have a tournament for a game
like Battle Cry which has unbalanced scenarios.
There are several well-known methods:
(1) Play matches (i.e. play each scenario twice, switching sides), which has
the disadvantage that each player must play the same scenario and opponent
twice, halving the variety of the tournament or doubling its length.
(2) Only play the most balanced scenarios. Disadvantage = the need to be
sure that the scenarios really are balanced (gather lots of data beforehand,
and trust its accuracy), and this limits which scenarios can be played.
(3) Have players bid victory points to pick which side they'll play.
Disadvantage = all players need to be quite familiar with the scenarios to
bid intelligently.
So we came up with a 4th scheme I've not seen before.
(4) Divide players into 2 groups (call them group A and B). Play several
rounds; each round pair off an A player vs a B player. The same scenario is
played by all players in a round, and all the A players play the same side
in a given round. E.g. in round 1, all the A players are CSA in Antietam.
Next round, play a new scenario, and shift all the A players down one seat
so they face a new B opponent.
E.g.
A1 plays CSA Antietam vs B1
A2 plays CSA Antietam vs B2
A3 plays CSA Antietam vs B3
next round
A3 plays US Shiloh vs B1
A1 plays US Shiloh vs B2
A2 plays US Shiloh vs B3
and so on.
Thus you get maximum variety: each player will never need to play the same
scenario or opponent more than once.
A player's tournament score is simply the total number of flags they killed
in all their games. All the A players are competing (in the tournament
sense) against all the other A players, and all the B players are competing
against all the other A players. Thus it doesn't matter if the scenarios
are balanced or not!
After whatever number of rounds, pick the highest-scoring A player and the
highest scoring B player. They go on to play a final round via a match (or
some other "balanced" way). If there is a tie in either group, you could
send them all up. (For this reason you might want to ensure that unbalanced
scenarios get distributed both ways. If the A group scores range from, e.g.
4 to 12, while the B group scores range from 16 to 24, then ties would be
more likely in the A group and thus unfairly give more A people a shot at
the final round...) Alternatively, you could simply count victories and
ignore the margin of victory, so a players score would always range from 0
to n in an n-round tournament.
Hs anyone played a tournament like this, or know more about subtle details
of running such a tournament? References or URLs are welcome...
The only obvious disadvantage is the need for a final round that not
everyone gets to play in, although many people consider that an advantage in
a tournament since it provides an element of drama.
A subtle disadvantage is that if the 2 best players happen to both be
assigned to group A, then one of them will lose out to the best player from
group B who might really be weaker than all the group A players but still
end up with 2nd place in the tournament. A way to handle this might be to
do a multiple elimination process, where after some number of rounds, the
top half from each group progresses to the next round, where the sets are
randomly chosen again.
Another subtle disadvantage is that if one of the players in group A is much
stronger than the others in group A, then the group B players who happen to
get paired with him will be unfairly penalized in their scores vis-a-vis the
other group B players who didn't play this person. So ideally you'd play
round-robin (each group A player paired against each group B player).
How thorough/careful/accurate you can get obviously depends on how many
games can be played, i.e. how long the tournament can last.
We got the basic idea of this from the Duplicate Go tournament at this
year's Go Congress in Denver, where all the weaker players competed against
each other in the tournament by playing out identical go endgames against
all the stronger white players (who were also competing against each other
in the tournament). Each black player faced off against each white player
round-robin style (i.e. there were n rounds, and 2*n players total in the
tournament). The 2 groups were simply separate divisions, each with a
winner, rather than having a final round.
russ
http://www.kofightclub.com