[pigpenz]

I am curious if people would be willling to spend the time to play in a full 26 board 13 table game...mostly cause of the time committment...thanks...and please post any comments you have , I would be willing to run one game a week if there is enough interest.

[Elianna]

I voted no. I would be interested if it were some sort of fast pairs, but there's just SO MUCH waiting time in between rounds for me that it makes the thought of playing in longer online tournaments agonizing. I think that if two board rounds, five minutes a board, and three or more boards four (even three!) minutes a board would be about the times that would tempt me. I realize that other people might think that this is too fast (or too slow?), so this may not be feasible, hence my voting no.

[omeroj]

Me too vote no, i prefer tourn speedy and fun:)

Omero

[hrothgar]

This might seem like a non-sequiteur, however...

I would very much like to see BBO implement some type of fully meshed tournament. If said tournaments were implemented, then I would be interested in playing in "long" events... In the absence of fully meshed tournaments, I don't see nay advnatage in playing in one 26 board event as opposed to a pair for 12 board events...

[Elianna]

what do you mean by "meshed"?

[hrothgar]

Elianna, on Jun 22 2005, 10:41 PM, said:

what do you mean by "meshed"?

A fully-meshed tournament is one in which every pair plays an equal number of boards against every other pair... Howell movements with 4,5, or 7 tables are examples of fully meshed movements.

Currently BBO implements movements that are primarily designed to be stable/robust. The movements are very forgiving and easily adapt to pairs dropping out of the event. However, thes movements aren't balanced and the events are pretty much crapshoots...

[hrothgar]

Jlall, on Jun 22 2005, 11:37 PM, said:

hrothgar, on Jun 22 2005, 03:00 PM, said:

Elianna, on Jun 22 2005, 10:41 PM, said:

what do you mean by "meshed"?

A fully-meshed tournament is one in which every pair plays an equal number of boards against every other pair... Howell movements with 4,5, or 7 tables are examples of fully meshed movements.

Currently BBO implements movements that are primarily designed to be stable/robust. The movements are very forgiving and easily adapt to pairs dropping out of the event. However, thes movements aren't balanced and the events are pretty much crapshoots...

even in the finals of the Blue Ribbon and Life Master pairs, the most prestigious pair games at NABCs, the movement is not "meshed." I guess those are a crap shoot too.

Not comparable...

Formally, bridge tournaments can be modeled as statistical sampling problems. In an ideally world, we wouldn't have any such thing as time constraints. Each and every pair would play a large number of boards against each and every other pair. We would then me able to identify a winner with a high degree of certainly. In reality, we don't have nearly enough time to implement any such scheme. Accordingly its necessary to make compromises. Howell movements are one such compromise. Field reduction - eliminating those pairs who score in the bottom half of the field during day “X” of the trial – is another such compromise. With this said and done, the organizers of these events go to a lot of effort to make sure that they are able to produce accurate results

As I noted earlier, the “pair matching” systems on BBO seem to be optimized to produce “robust” movements. BBO has self-healing movements that can dynamically adapt to the loss of pairs during the course of play. However, these features significantly decrease the accuracy of the results. Please note, the administrators who run the Blue Ribbon pairs don't need to worry about significantly numbers of players quitting in mid-event. Different playing environments lead to different design choices...

For what its worth, there are some “pair matching” algorithms that are able to combine robust performance with a fair degree of statistical accuracy. So-called barometer events in which pairs a seeded based on performance are quite good at identifying the “best” and “worst” pairs at the expense of less accurate rankings in the center of the distribution... Even so, the accuracy of barometer events is very much a function of the length of tournament.

[pigpenz]

thats why i was setting 26 boards 13 tables as max....every EW pair plays Every NS pair. Time is the problem generally 15 min round no smoking breaks I have noticed that the way the acbl games are ran, they have sections but still your are competing agaisnt everyone overall for yur matchpoints or imps....so sometimes 45% at matchpoints may place or -imps may place.

[JanTucson]

I heartily agree with Elianna's first comment. The wait between rounds IS interminable !

[fifee]

The wait is long between rounds but we can always spend time reading email, studying or system notes

Playing from home makes it hard to find a 2 to 3 hour uninterrupted time for a tournament for me.

[MickyB]

26 boards sounds too long to me. Might be ok with my regular pard if it happened to start at the right time, but could be intolerable with a pick-up.

I disagree on the need for a "meshed" movement - it is entirely possible to have a fair movement which doesn't involve you playing everyone, or an unfair movement that does make you face each other pair.

[uday]

Unclocked games offer much smaller wait times between rounds. The price paid is that there are playbacks, and that we cannot produce NS/EW winners (only overall winners), and that the length of the tourney is extended. Additionally, unclocked tables with a missing player will just sit there, waiting (as opposed to clocked, which move along when the round completes)


It has been my experience that while a clocked 12 bd pair game runs for 1.5 hours, an unclocked runs for 1 hour (for the fast players) to 2 hours (for the slow players).


So, if event duration is an issue for some players, perhaps you could consider using an unclocked pair movement for the longer events.

Another possibility is to run a survivor movement. This is kinder to long events since players can bail at the end of a round (the movement compensates by tossing the lowest scoring pair as well) whenever they get tired. And including a cut makes sure that the people who're getting chopped up are excused.

[helene_t]

In an indy (I guess this may be somewhat off-topic), the delays can be reduced to one-third by playing all three constelations at the same table in succession.

[Fluffy]

I would love to play that, but I wouldn't want to wait between switching rounds.

Unclocked solves it in some way, but I would love some new kind of competition....
4 minute clocked tourney (you get 0% if you let the clock go on), or a clock survivor tourney, where the 2% of pairs who finnish last are simply booted .

Of course that nonsense is just because I am normally much faster than others and would like to get advantage of it.

[joker_gib]

Fluffy, on Jun 23 2005, 02:07 PM, said:

I would love to play that, but I wouldn't want to wait between switching rounds.

Unclocked solves it in some way, but I would love some new kind of competition....
4 minute clocked tourney (you get 0% if you let the clock go on), or a clock survivor tourney, where the 2% of pairs who finnish last are simply booted .

Of course that nonsense is just because I am normally much faster than others and would like to get advantage of it.

That would be interesting but, unfortunately, I know that a lot of players would let the clock go as soon as they have a bad result !

That's quite frustrating !

Alain

[Al_U_Card]

While we are into novel (bizarre?) suggestions, how about the round is clocked (like fast pairs) and your score is adjusted based on when you finish (the avg. finish time remains unchanged while the fastest get bumped up and the slowest get bumped down.....maximum adjustment say, 1 board per 10 played?

[dogsbreath]

Hi ..
although i voted 'for' .. it will still likely be tiresome to play in. In team matches, where speed of play is generally much faster, 16-board matches often run into problems (leavers etc)

8 mins per board is still (imo) much too slow for on-line play and all tourneys seem to run at the convenience of the slowest players (complicated by subbing problems).

Rgds Dog

[pigpenz]

the problem in the past is sometimes people will play the clock when they know they are going for a zero, so they try to give their opps who are headed for a good score an avg-- by not finishing the hand, now this makes the director have to do alot of work always adjusting the score, it would be nice if you could assign a zero for the offenders and the real score for the nonoffenders, but i believe the way the score adjsutment is now you cant assign that type of penalty

[Al_U_Card]

pigpenz, on Jun 23 2005, 10:29 AM, said:

the problem in the past is sometimes people will play the clock when they know they are going for a zero, so they try to give their opps who are headed for a good score an avg-- by not finishing the hand, now this makes the director have to do alot of work always adjusting the score, it would be nice if you could assign a zero for the offenders and the real score for the nonoffenders, but i believe the way the score adjsutment is now you cant assign that type of penalty

Well, thanks to Fred and the marvels of modern computing, I am sure that there could be a "running clock" on each pair's bids and plays. This is where the total time would show as 8 minutes for the hand, 7 minutes 30 sec. for E-W and 30 sec. for N-S, and you know who gets the penalty assessed.