Future BBF events discussion

[frank0]

Summarize all the issues arise from event 6 so far

 

-The big group RR structure makes it difficult to fairly deal with people whose opponents drop out.

-The current rules dealing with drop out/unfinished match opens up the possibility of exploiting the rules.

-Determining the seed: Only use the score vs. top half finisher (suggested by ovncylmz).

-# of entrants already near the max. a few human can deal with.

 

Possibilities for the future events

-Update the related rule for unfinished matches.

-Smaller group even for large number of entrants.

-Start with KO even for the first stage (New problem, the seeding method we have now is quite inaccurate).

-Possibly have BBO "month long" tournament to replace this eventually.

[gwnn]

Easiest way to handle the comparisons is to have a number of groups that are a power of 2 so that you can have a fixed # of people advancing. For 55 entrants, you can have 8 groups (7x7+1x6).

[frank0]

Easiest way to handle the comparisons is to have a number of groups that are a power of 2 so that you can have a fixed # of people advancing. For 55 entrants, you can have 8 groups (7x7+1x6).

In this example isn't it too big of an advantage for players in group of 6?

[gwnn]

I don't think so (here, it would be ~33% instead of ~29%, assuming two people progress). This will happen in any case in which you have awkward numbers of participants. And comparing the performance of people from different groups is anyway a very questionable procedure.

[gwnn]

PS I don't think many people really care about the seeding. I may be extrapolating from N=1 though. The FA Cup is one of the most prestigious soccer events and it has no seeding at all (once the field is similar to what we are likely to see in the BBF challenge events).

[Phil]

PS I don't think many people really care about the seeding. I may be extrapolating from N=1 though. The FA Cup is one of the most prestigious soccer events and it has no seeding at all (once the field is similar to what we are likely to see in the BBF challenge events).

 

I do - at least to balance the groups.

[Phil]

With matchpoints, you probably don't need 4 x 16. I'd say 4 x 10 is sufficient.

[m1cha]

-Start with KO even for the first stage

I prefer the group stage: Play against several opponents rather than only one and then possibly drop out.

 

-Possibly have BBO "month long" tournament to replace this eventually.

That sounds like something completely different.

[smerriman]

That sounds like something completely different.

See here for prior discussion: http://www.bridgebase.com/forums/topic/75830-bbf-challenge-event-6-format-poll/page__view__findpost__p__909257

 

Fundamentally it's the same thing, just that you get to challenge everyone simultaneously with the same hands rather than only your group, and since it's the same hands you can play many more in the same timeframe, so more accurate results.

[barmar]

Another nice thing about using a monthlong for the qualifier is that it's easy to just ignore players who don't complete it, they're automatically disqualified and don't disturb the seeding.

[m1cha]

Thanks for the link. I missed that discussion.

 

Fundamentally it's the same thing, just that you get to challenge everyone simultaneously with the same hands rather than only your group,

That would not be the same thing, it would rather be a kind of survivor tournament running over several days per stage. I like the survivors but I prefer the challenges by a wide margin. To me the particular charm of the challenges is in the personal touch, playing against people you would not ask otherwise.

 

And if it would mean playing 100 boards in 3 days at a given time, I think I would not be interested. Okay, so you probably think of extending the period of a challenge from 3 days to 2 weeks, that's fine. But if you do that, why not program a bot to issue the individual challenges to the players at the start of the group stage, auto-accepted and automatically limited to the end of the group stage. That would remove the nuisance of checking, cancelling, rejecting and re-issuing of challenges. The bots could even collect the scores automatically to a log file.

 

On the downside of this, we would play some boards and get the result two weeks later after we have forgotten the hands. This is another thing that may turn out to be uninteresting. Perhaps the concept of challenges running for 3 days is not that bad at all.

 

and since it's the same hands you can play many more in the same timeframe, so more accurate results.

Yes but if we all play the same hands, couldn't someone sign in under several identities using one to check the boards and the other(s) to play them with open cards?

[m1cha]

Another nice thing about using a monthlong for the qualifier is that it's easy to just ignore players who don't complete it, they're automatically disqualified and don't disturb the seeding.

Yes, that was a problem this time because several players signed up mistakenly. I've got an idea how to adress this. I'm gonna make a suggestion, perhaps tomorrow.

[smerriman]

That would not be the same thing, it would rather be a kind of survivor tournament running over several days per stage. I like the survivors but I prefer the challenges by a wide margin. To me the particular charm of the challenges is in the personal touch, playing against people you would not ask otherwise.

The idea was that you'd still need to sign up via the forum to be able to access the tournament (as opposed to a free for all like the Survivor/real daylong tournaments), thus keeping the personal forum touch, and basically keeping cheating out of play.

 

You'd want to limit forum signups to people actually active on the forum (a few of the non-responders in the current event had 0 posts, for example) - this should really be the case for the challenge events regardless. If someone has multiple active forum accounts with multiple posts under each and really wants to cheat in a friendly forum tournament..

 

On the downside of this, we would play some boards and get the result two weeks later after we have forgotten the hands. This is another thing that may turn out to be uninteresting.

 

Yes, that's the only real problem. It would be nice if the scores could update in real time somehow, but that may be beyond BBO's capability at the moment.

[sfi]

Yes, that was a problem this time because several players signed up mistakenly. I've got an idea how to adress this. I'm gonna make a suggestion, perhaps tomorrow.

 

Asking people to reply to the thread rather than simply responding to a poll seems to lead to a much higher participation rate. It's a small demonstration of commitment, but could be enough of one for this purpose.

[m1cha]

Asking people to reply to the thread rather than simply responding to a poll seems to lead to a much higher participation rate. It's a small demonstration of commitment, but could be enough of one for this purpose.

Yes but I think it's not about the commitment, it's rather about understanding the event. When it turned out a week ago that some players did not start playing, I identified 6 people who had not played at all and who were new to the forum (0 posts, first login within a few months). I sent them a message, got a reply from 4 of them and it turned out they all were not aware that they had signed up for the event! They must have clicked Yes to "Play?" in the vote, obviously, but what exactly does it mean? After all the forum was linked in the news-feed and must have got lots of visitors, some of them just make mistakes. In addition, post #1 was optimized for experienced participants but it is pretty long and half of the information is just confusing to new visitors while other information for new players is missing. I believe post #1 should be optimized for our uninformed visitors. The experienced people will always find what they want to know.

 

I believe that the organizers invented the sign-up-by-vote because it is very comfortable to experienced players - and to organizers also because, well, if you have to collect the participants from a total of 50+ posts, this will inevitably lead to errors sooner or later. I immediately liked the vote thing. But it should be made clearer to all visotors that this is the sign-up process and for what they are signing up. I believe if we do this we can minimize the dropouts in the next event.

 

So I have set up a template that can be used in the next challenge event(s). Here it is.

 

 

Cast your vote here to sign up for challenge event #0. (If you have signed up mistakenly, please click the "Delete My Vote" button under this vote.)

O Yes, sign me up. I have read post #1. I know how to send challenges to my opponents, and I will take part in this forum on a regular basis.

O No, I will not play this time. (You may click here if you think someone is interesed in knowing that you will not take part.)

 

[Delete my vote]

 

What is a challenge event?

A challenge event is an innovative form of bridge tournament in which the participants play BBO challenges sent to each other. After the start of the event there will be an announcement on this page about who you will play against. For the time frame of this event see below. The event is free of charge.

 

How to take part in this event

Before the event:

- Read this post completely.

- Make sure you have understood how to send and play challenges. (See this link if you don't know how challenges work.)

- Read the rule page. (See this link.)

- Find the score table at this link. Note the tabs at the bottom to scroll through the pages. (This may not work before the start of the tournament.)

- Make sure you have sufficient time to play the challenges while the event is running.

During this event:

- Play challenges sent to you.

- Send challenges to your other opponents. Make sure the challenges match the format of this event (for the format, see below).

- The players of a challenge must report the result to this forum (usually it's done by the last player to play the challenge).

- Check the score table occasionally.

- Keep track who you haven't played against yet and make sure you play all challenges in time.

- Keep up to date, read the forum at least once every 24 hours.

If you are ready to sign up, vote "yes" on top of this page.

 

The format of challenge event #0

- One challenge per player in your group during the group stage. (The group size has not been decided upon yet. It is typically between 6 and 12.)

- 10 boards per challenge during the group stage.

- Scoring is always Matchpoints (BAM).

- Best hand is OFF!

[KO stage format can be announced here later]

 

Time frame

- The group stage runs from ... to ... [fill in]

- The KO stage will start soon after the group stage has finished, the precise time will be announced here. [fill in later]

 

The groups

group 1: ...

group 2: ...

...

[This place will be filled up when the registration has ended.]

 

Challenge information and etiquette

Challenges can be played within a duration of 3 days from the time they were issued. If they are not accepted within 3 days, they expire. While they have not been accepted, they can be rejected or cancelled any time. Once they have been accepted, they must be played. A player who misses playing an accepted challenge has technically lost it, and a player who lost a challenge cannot proceed to the next stage. It is bad manner to accept and play a challenge shortly before the time of expiry because this would be putting pressure on the other player. This is a friendly event! Better reject the challenge and send it back. The same situation from the other side: When you have sent a challenge that remains unaccepted for a while, and if you think you may not be able to play it in the remaining time, you are urgently recommended to cancel the challenge! You may resend it immediately or later or wait for it to be sent to you. It is often a good idea to contact the other player by PN and make an agreement about when to play. When you get a challenge from someone you follow, it will be auto-accepted! This function helps people to play faster during the KO stage but please make sure that you don't run in trouble from auto-acceptance, play friends first.

 

Links

Link to explanation of challenges

Link to the rules

Link to the score

 

Further information

If you have questions about this event, feel free to ask by clicking "Reply" at the bottom of this post.

In case of trouble during this event please contact the forum as soon as possible. This is a friendly forum, people will try to help.

 

News ticker

[Announcements during the event may go here.]

This post should contain all information necessary to understand the event, but anything else might confuse beginners and should be kept off at least while the registration is running. In particular anything about seeds and tie breaks should be moved to the rules page.

 

I will gladly send this template by e-mail to interested parties.

 

The discussion is now open. :)

[barmar]

Maybe we could have a third choice in the poll, something like:

 

o I'm interested, but I don't understand how to play (post your questions by replying to this thread, you can change your vote later)

[gwnn]

Maybe gathering email addresses could be an idea, or making a BBO group?

[frank0]

Anyone thought about the seeding method of the first round of KO? It has not been accurate since event 3.

[nige1]

A KO with long matches scored at MP (BAM) is a good format but it has some drawbacks. For example, if the best player has an off-day, he can be eliminated in the first round.

 

Now consider a Swiss competition with BAM matches of the same length.

(Each round, you play somebody with a similar score, whom you have not yet played).

A Swiss is like a KO with repêchage.

You don't play everybody else but the matches are much longer than in a RR.

And opportunities for tactical dumping are rarer than in a RR + KO.

If a player drops out of a Swiss, it's not a disaster.

IMO, comparing formats, over the same time-interval, the end result of SWISS is a fairer reflection of relative skill than any other format.

[sfi]

IMO, comparing formats, over the same time-interval, the end result of SWISS is a fairer reflection of relative skill than any other format.

 

Maybe for the winner, and I'm not convinced even for that. However, the big problem is the unequal weightings for each round, and your final standing in a Swiss event depends enormously how lucky you are in your draw for the last round.

 

I've consistently stated that I would happily skip the first round of a Swiss tournament and sleep in. In exchange, the directors can assign me any number of victory points from 0-20. My claim is that the value of those victory points is essentially nothing and I would gain by the additional rest and lack of stress. Sadly the regs don't allow me to put this into practice - they remove whatever penalties are applied after doing the draw, so I don't gain the benefit of easier opponents.

 

A Swiss is dreadful for accurately sorting a field.

[nige1]

Maybe for the winner, and I'm not convinced even for that. However, the big problem is the unequal weightings for each round, and your final standing in a Swiss event depends enormously how lucky you are in your draw for the last round.

 

I've consistently stated that I would happily skip the first round of a Swiss tournament and sleep in. In exchange, the directors can assign me any number of victory points from 0-20. My claim is that the value of those victory points is essentially nothing and I would gain by the additional rest and lack of stress. Sadly the regs don't allow me to put this into practice - they remove whatever penalties are applied after doing the draw, so I don't gain the benefit of easier opponents.

 

A Swiss is dreadful for accurately sorting a field.

IMO, in sfi's thought experiment,

 

1. A simulation would prove that your expectation of winning overall would be greater after winning rather than after losing the first-round.
2. The ranking order in a Swiss is likely to correlate better with "God's ranking", than in competitions with other scoring-formats.

For a BAM Swiss pairs, I advocate a simple scoring system. e.g. for each 64-board match you score winning boards + 50 (say) if you win. (i.e. winning or losing is more important than the winning-margin).

 

The popular and prestigious SBU Winter 4s and EBU Spring 4s rely on a format called double elimination -- similar to a knock-out -- but you continue until you are beaten twice. IMO, a Swiss is similar, except that you aren't eliminated.

 

Incidentally,

CONGRATULATIONS TO PAUL GIPSON,

 

who won the SBU Winter 4s last weekend, against formidable opposition!