Always a Titanic Struggle ...
This year there were several notable statistical flukes.
The first was that this year, for the second time in the past
11 years, Steve Koleszar both won Titan and didn't pick
it as his team event. (Steve is the first two-ime Titan
winner.) The other nine years he picked it as his team event,
but, although often doing quite well, never won in those years.
There was a second statistical fluke involving Steve this
year. In his semi-final game, Andrew Gross (this year's two-player
Titan champ) attacked Steve with a titan (10-4), two serpents,
two angels, two gorgons, with an angel to summon. Steve was defending
in a tower with titan (8-4), two guardians, a ranger, and an
angel. Andrew was a huge favorite, but the dice gods were
smiling on Steve, and Andrew's chance to be the first person
to win both events in the same year ended.
The third statistical fluke was that out of 51 official games,
six ended with two of the last three players remaining mutually
eliminating each other to leave a winner by default who didn't
get the chance to personally defeat his opponents. Normally this
only happens once or twice through the course of the event.
I changed a few procedural things from last year and want
to report on the results.
One thing I have
done over the years is try to make the multi-player tournament
suitable for a variety of people, including those that just wanted
to play drop-in games, play lots of Titan or just play
to win. This year I made a change to allow people who wanted
to play a lot of Titan, to get a better chance to do so.
The change was to allow players to play as an eliminator after
playing six games that counted. Several players took advantage
of that. The most notable player to do so was Dan Strock who
played in 11 (six normal, three as eliminator, semi-finals and
finals) official games, several official two-player games, and
several unofficial games that I noticed (including teaching new
recruits). He deserves the Titan Marathon Player award for extensive
play this year.
After one of the semi-final games ran over the (generous)
time allotment last year and how the game ended didn't make anyone
involved happy, I made of a couple of changes. One was to institute
a rule for using timed turns if a semi-final or final game was
running long. The other was to use an objective formula for producing
a winner if, despite time controls, a semi-final or final game
went over the time limit. However, this year none of those games
went anywhere near their time limits. The time limit on the final
game was new, but I felt that there needed to be some limit,
as I didn't want someone to have to quit because their plane
was leaving on Sunday. So, I used a cutoff of 6am Sunday, which
gave 12 hours to play the finals (which is about three times
the normal time needed).
Another change was in response to a suggestion that players
start fresh in the semi-finals. Previously, in the case of a
mutual score between the last two players in a semi-final or
final game, the higher seed would advance. This year I instituted
a roll off between any players that mutualled in the semi-finals
or finals. However, no mutuals occurred this year.
Even though it didn't come into play, people liked the roll
off method of settling mutuals. Next year I will be using this
rule for preliminary play, as well as for the semi-finals and
finals.
One big topic was seeding. I have been using an objective
formula for seeding because it is hard to fairly place everyone
for their semi-final matches. Two problems with this system have
been pointed out. One is that it may (did) cause players to play
extra preliminary games to modify who got to play who in the
semi-finals, even though they were sure to be in the semi-finals
and would have preferred to have done something else with that
time. The other thing it didn't do was separate people who didn't
want to play each other (typically people from the same geographic
area).
The discussion was very fruitful, and already I almost have
a complete system set up for next year. The preliminary results
will still be used to decide who plays in the semi-finals, and
act as a tie breaker between semi-final participants who don't
make the finals and finish in the same position in their semi-final
game. However, the seeding will be done by random draw with some
extra constraints.
The extra constraints are a system that might be useful for
other people running multi-player events. When a player registers
for Titan, they will be allowed to name a group of up
to four players (including themselves) that don't wish to meet
each other in the semi-finals. These other players must also
list the same group when they sign up to play. This allows groups
of up to four players to be split up if they all make the semi-finals.
(If, when signing up, players list inconsistent groups, there
is a simple system for resolving conflicts in a fair manner.)
Some people were concerned that a group of four "good"
players might try to take advantage of this to arrange not to
play each other. However, since the whole idea of seeding was
to split up the "good" players, this seemed like it
could only be a good thing. (Especially since the current seeding
system wasn't doing that great of a job at this.) We decided
to not force people that are part of a team to take advantage
of this system. I will be prohibiting teammates of a player with
Titan as their team event from playing in the same semi-final
game. This won't break the splitting up of groups, but I do have
a little work to do with figuring out how to select the match-ups
randomly. I would like each possible match-up to be equally likely.
In spite of the success of this year's event, I am looking
forward to an even better one next year!
|