[glen]

What about iphone speedballs - what is the speedball's chance in cell?

[NickRW]

hrothgar, on Aug 4 2008, 02:33 PM, said:

For what it’s worth, I’ve spent a fair amount of time speculating....<snip>

One thing is for sure, if people have to put money into it you'll get far fewer apparently interested people, but those who actually sign up will be much more likely to really be interested.

I don't see why either BBO or the TD shouldn't be able to benefit if there is money involved. BBO provides the infrastructure and the TD, if he/she is any good at the job, make sure there is decent game and give of their time. In the UK at any rate, the training to become an approved TD is reportedly not cheap either.

I know this is yet another idea for Fred, Uday and company to implement, but it would be easy enough I suppose for BBO to maintain a record of those who don't complete competitions. Then, instead of TDs having to maintain long lists of enemies (which are reportedly expensive in terms of performance for the server to check) that are not to be admitted, the TD could simply tick a box when setting up the competition - "nobody with bad record to be allowed in". Bad points against your name could be made to decay over time so you wouldn't stay banned forever.

Another idea for good competition where much of the effort in organising it devolves to the players themselves is team league/ladder competitions. You need someone prepared to be an organiser/secretary and, I guess if you're serious enough, someone from each team prepared to serve as member of an appeals committee in case of argument that captains can't resolve.

Another idea for some slightly serious, but casual competition is for BBO to deal a fixed number of boards per 24 hour period - say about 60. You can play each board only once in the day. You can move to another table if you want. You can change partners if you want. You can play as many or as few from the set as you want. You can play as quick or as slow as you want. At the end of the day you get an email giving you the results on the boards you played with a percentage MP score for each and an average of the averages. The top 3 or so players for the day get their name mentioned, but no other prize or masterpoints.

I don't know how much, if any, of this goes on already, not having sampled the delights available as yet (doesn't sound like much from the comments on the forums).

Nick

[hrothgar]

NickRW, on Aug 4 2008, 07:37 PM, said:

I don't see why either BBO or the TD shouldn't be able to benefit if there is money involved.  BBO provides the infrastructure and the TD, if he/she is any good at the job, make sure there is decent game and give of their time.  In the UK at any rate, the training to become an approved TD is reportedly not cheap either.

BBO is in charge of providing/maintaining the infrastructure.

If BBO wanted, the could knock any user they wanted offline and preventing them from connecting back to the tournament.

I don't believe for a moment that Fred / Uday would gimmick the server in order to collect the occasional bond and improve the bottom line. However, I have absolutely no doubt that folks would start complaining that BBO was doing just that.

The best way to sidestep these types of issues is to create a system where BBO doesn't collect any kind of $$$ from a bond forfeit.

I'd also argue that the bond is designed to deter players from disconnecting. When a player disconnects from a tournament this reflects either an Act of God or a failure on the part of the TD / infrastructure provider. I don't think that its reasonable to reward these entities with extra income. Instead, I prefer to see the bond distributed to the folks paying the tournament fee,

[vuroth]

Interesting.

I'm not an ACBL member at the moment. I lapsed when my bridge partner and I started having kids (not with each other!!) I don't plan on rejoining until I have improved, and have a regular partner again.

That said, I would very much have considered using BBO to build up my masterpoint total (yes, I know they're not necessarily meaningful). From the sounds of it, for someone like me with <20 MP, finding a suitable ACBL game is going to be difficult?

[Gerben42]

Quote

Any interest in super speedballs, like 4 minutes per board? I think that even the speedballs are slower than they need to be.

You did say 4 minutes for ONE board, right? And that's supposed to be SUPER? Now 4 minutes for 2 boards, that's Super Speed

[barmar]

NickRW, on Aug 4 2008, 12:37 PM, said:

I know this is yet another idea for Fred, Uday and company to implement, but it would be easy enough I suppose for BBO to maintain a record of those who don't complete competitions. Then, instead of TDs having to maintain long lists of enemies (which are reportedly expensive in terms of performance for the server to check) that are not to be admitted, the TD could simply tick a box when setting up the competition - "nobody with bad record to be allowed in". Bad points against your name could be made to decay over time so you wouldn't stay banned forever.

They already track bailers in the MBC, and automatically ban them temporarily. This seems like a reasonable extension of that.

[rigour6]

I think the major reason speedball dominates is that is scheduled essentially on top of the other tournament. If they were spaced an hour apart they wouldn't be competing for the same pool of players.

I end up in speedball generally because a) it's lunch so I don't have time but more often when I check the regular, it has 1 entry, and when I check speedball starting a few minutes later, it already has 8. So I go with the larger field.

[matmat]

rigour6, on Aug 6 2008, 11:13 AM, said:

I think the major reason speedball dominates is that is scheduled essentially on top of the other tournament. If they were spaced an hour apart they wouldn't be competing for the same pool of players.

This is an interesting point. Professional sports actually worry about having too many teams in their leagues and thinning out talent. When there is a GLUT of tournaments it is hard to get a good one going.


Also, I think that possibly one way of curbing people jumping team games and tournaments is something like a 5-5 rule -- if you're not back within 5 minutes of disconnecting, i.e. in time to finish the board/round/whatever, you're automatically banned for 5 hours.

[kenberg]

It's exactly this "the regular has only one entry" that I hope to cope with. Success breed success, failure breeds failure. Getting a partner to play oin the 1 edt regular game now seems pointless to me, since it often doesn't run. But of course then I am not listed among the entries and others see a small number of potential players.

I'm not searching for any exotic solution although several ideas sound promising. I think I'll try to watch and see which of the regular games often run, try to fit my schedule to them, and let the other games fend for themselves. Not ideal, but I don't really see much else.

btw I looked into unibridge. Closed to new members it says.

In some ways I might prefer 4 minute boards to 5 minute boards. If I am going speedy, I may as well go the whole distance. But really I prefer regular. It takes time to find a really idiotic line of play.

[TimG]

kenberg, on Aug 6 2008, 12:43 PM, said:

btw I looked into unibridge. Closed to new members it says.

A friend invited me to play in unibridge game a few weeks ago, I wasn't a member, but an e-mail from the friend to the director/organizer got me into the game without any problem.

As I mentioned before, I found it agonizingly slow. 21 minutes for 3 boards is really a lot for an online game. Considering that 21 minutes for 3 boards is typical for face-to-face and you've not got any record keeping, hand sorting, board changing to do online, the time is actually more than allowed in typical face-to-face play. It wasn't that the play was slow, but that there was typically 8-9 minutes to wait between rounds.

My opinion is that it's not the fact that there are few people signed up for the ACBL games that is killing the game, but that people aren't signing up because there is very little demand for play under the time conditions. That doesn't help you, of course, and I truly do wish you luck in finding the type of game you are looking for, I'm just not optimistic.

[rigour6]

Let me give you an occasional TD's perspective:

1. Unlike a ftf game, sitting at a table waiting for the movement to be called does not give much opportunity for socializing. As a result, it's painful.

2. I wrestled with this problem a lot in my early TD days. You set the clock low, people play the clock when they see they're in a bad board, that leads to adjustments, which take time you may not have. You set the clock high and the tournament draggggggggs along.

3. My answer may seem counter-intuitive but here's what I did:
i. I turned the clock OFF.
ii. I set the timer for 6 minutes a board.
iii. I ensure all my tournaments have no more than 3 rounds. 3 rounds of 3 or three rounds of 4, that's it.

What happens is these tournaments actually run FASTER than timed tournaments. Here's how it works, I've got 10 tables and that's about their speed. Tables 1 and 2 finish, whammo, they switch tables and move on to round 2. Everybody else is playing round 1. Then tables 3 and 4 finish, and off they zoom. Each table as it finishes, only has to wait until another table finishes and off they go to the next round. Finally only the slowest two tables remain, now this is typically more a product of disconnects than actual play but anyway, off they go.

Now everyone is in Round 2 and the field has been "seeded" by "speed of play" rather than score. The fast guys are playing other fast guys. The slow guys are playing slow guys. In Round 2 if you go from fast to slow, maybe the third round you're in a slightly slower group, or maybe you were the fourth table to finish round 1 but are the 2nd to finish round 2, so you move up into the fastest group.

The cheetahs blow through the boards like a hurricane, no waiting. I've see them finish Board 9 before some people started Board 6 pretty regularly.

OK, so what's left: 2 things. One is the 6 minute timer, even in an unclocked tourney, will auto-pull a board if you don't get to it in time. You have for example 18 minute to get 3 boards in. If you get to Board 3 with 4 minutes left, fine. The clock will go negative, reminding you to speed up, but it won't pull the board you're playing. You absolutely cannot avoid a bad score through slow play, a tactic which is the bane of the clocked tournament and the suspicion of which causes apprehension and tension between players. If you start a board, you'll finish it. Period. But if get to the third board and the timer says you have 2 minutes, whoops, what's this? The computer auto-pulls the board and assigns an average (not ave minus, just average) and this tends to catch you up with the group, as you're soon off to the next round.

Now we get to the end of the tourney and we're waiting painfully for the last 5 tables to finish. And I have an answer to that, too. I now reset the timer to 4 minutes. Which typically means the last board for them gets pulled, there's no way they can get to it in time. They auto-average the last board, people aren't sitting around forever watching the turtles play, and the people that lost the last board? Well, they were too slow and here's the final key point: they knew this would happen going in, because I announced it in the rules of the tourney.

This creates a "culture of play" in which you play at the speed you are comfortable. It's not a race, the vast majority of players don't have a board pulled. But it's advertised as a fast tournament and if you don't meet that description well either accept you play one board fewer or find another tournament.

My latest tourney (today) 3 rounds x 3 Boards each:

Start:
+3 first Board finished by fastest table
+7 2nd board finished by fastest table
+9 first board finished by slowest table
+10 2 tables finish Board 3 (one of them passed it out). They could now start an 8 minute wait in clocked tourney. Instead, they immediately begin the next round.
+ 17 Fastest Tables finish Board 5. No, I'm not making this up.
+18 2nd board finished by slowest table. The system now tries valiantly to get them back into the flow by pulling Board 3, averaging it, and they too are now off to Round 2, actually 1 minute ahead of some others.
+19 Everyone is in Round 2. (Boards 4 to 6)
+32 Slowest table finishes Board 5. System skips Board 6 for them throws them into Round 3 (board 7), 6 minutes ahead of the last group to finish Board 6.

+37 Everyone is in Round 3 now. Oh, except for the (get this) 4 tables who have already finished the tournament. One of them did it 5 minutes ago. That's right, 9 boards played in 32 minutes. These people now have 21 minutes they can vacuum the flat, read a book, or play somewhere else. By rights, they should take a minute of their time and say a prayer for my continued good health.

+ 53 Last tables finishes Board 9, tournament ends.

29 tables x 9 Boards = 261 Boards Played.

Boards auto-averaged by the system: 3 , or 1.2%
(I just chose this tourney as the last one I TD'ed, so this may be lower than usual because even though I pulled the timer down in the last round, they all made it through Board 9 - in part because the 3 most likely to be slow had received those earlier boosts at Board 3 or 6.).


I set the tourney parameters such that 3 boards which would have been played by the slowest of 29 tables didn't get played. 52 teams played 9 boards and 6 played only 8. No-one played 7. The slowest of the tables managed 9 boards in 53 minutes, in other words less than 6 minutes a board. Nobody ran the clock to avoid a bad result, I needed to make zero clock-based adjustments.

If you set the parameters like this, people can enjoy a fast game and if they fall behind, worst thing that happens is they play one fewer board, for which they take an average.

[jtfanclub]

Unclocked tourneys are nice for fun games, but don't work for the big events. Not if you can figure out whether the cheetah section or the slow section tends to be the better guys, and then put yourself in the other half deliberately.

[rigour6]

While I take that point, let's be realistic. What's a BIG tourney online, 12 boards? Maximum number of rounds 6.

You're in a "serious" tournament with let's say 40 tables. 80 pairs. So overall, you'll play 2 boards each against 7.5% of the field. Of the 6 teams you play, they'll at best have one other pair they'll play in the round that you will also play. Probably 80% of the pairs in this "serious" tournament (too lazy to do the math) will reach the end of the tournament having played neither you nor anyone you played against.

Two nights ago at my local club, I played 24 boards and at the end of the night, even with a half table sit-out, I'd played 3 boards each against 80% of the field, and the 2 teams I didn't play had played either 100% or at the very lowest 87.5% of their games against common opponents.

I don't want to misuse statistics here. The real determinant of the accuracy of the results is based on the number of teams entered and the number of rounds. But you can see my point: a large online tournament of 6 rounds maximum will never reach the threshold of being statistically very accurate. As proof of that, I've won a couple.

I'm not trying to make the case to stop holding "serious" tournaments online, nor do I want to fall into the trap of saying, well it doesn't matter if you play 7.5% of the field or in my case half that. But I'm not sure the difference justifies the much greater time of the players which a clocked tournament requires - particularly if the risk is that a very strong pair will deliberately slow their play in the hopes that will give them weaker opponents in the following round. Even if you accept the underlying hypothesis which is that the slower you play, the worse bridge player you are, can better players really help it? Are they prepared to deliberately enjoy the game less just to marginally increase their chance of winning? My guess: if there's money on the table, mmmmmmaybe. Which is of course another reason I'm no fan of how money prizes distorts the event.

[blackshoe]

The purpose of duplicate is to determine through direct score comparison which contestant is the best. A "duplicate" where the number of comparisons is small, whatever the reason, defeats the purpose. Online "duplicate" has, almost invariably as I understand it, a small number of comparisons. There's no point. Might as well play rubber or Chicago (can the BBO software do Chicago?) Or, perhaps, head to head teams, although even there I don't think 12 boards is enough.

[hrothgar]

A few years back I made what I thought was a relatively simple recommendation:

Masterpoint awards should reflect the statistical properties of a tournament.

Its relatively easy to study tournament designs and evaluate how effective a given format is at ranking pairs. Some formats (a single eliminate KO with long matches) are quiter good at identifying the best pair, but quite poor at ranking pairs 2/3/4/5 whatever. Other formats (a fully meshed pairs movement) are much better identifying the top "N" pairs, but not as good at identifying the very best pair. Some other formats (say, 12 board speedballs with 100 pairs) are complete crap shoots.

In theory, one could weight masterpoint assignments based on the statistical properties of a given format.

A single elimination KO should ahve a very big payoff for the top team. However, the masterpoint assignment for teams 2-X should be pretty flat. (One could even argue that any team eliminated by the top team should share in the masterpoints)

I doubt that this type of proposal would go very far. (Selling masterpoints seems to be incompatible with designing accurate ranking systems). However, I think that the idea has some merit.

[rigour6]

I could definitely agree with that. I think we have to recognize, however, that masterpoints, at least as the ACBL hands them out, only very indirectly reward or measure skill. Their primary purpose is as a marketing tool for people to maintain a membership in the ACBL and an interest in continuing to play, as the act as a type of measure of progress, in the same way an odometer measures progress. As a marketing tool, masterpoints are quite successful, and for this reason I think they are a good idea.
They add to enjoyment of the game, as we've all seen someone get motivated as a masterpoint milestone approaches.

Masterpoint rewards for online tournaments are imnsho ridiculously high, but that's part of a larger masterpoint inflation issue, and perhaps its tempered by the fact that the points online are "colourless". But just to give an example: I was recently privileged enough to be on a team which made it to the national team championships in Montreal, and I think our record there in the final field of 14 B teams from across the country was about 50-50, we came 6th overall. 3 days of play, 13x12 board matches, $500 entry fee or some such. At the end of it all, I can't recall now, something like 4 masterpoints. For a fraction of fees and effort I dare say I should be able to generate many times that if I took three days and played in all the online tourneys I could. So in my own case, I just recognize masterpoints for what they are, a sort of bridge frequent flyer miles which say more about how experienced a bridge player than most anything else. And nothing wrong with that.

The ACBL is of course aware of the "problems" their present creates in terms of ranking teams by skill, and they're working on some new system to deal with that. They will want to preserve the membership retention advantages of the present system but also create a system which reflects how well you've played in top events over the past year for example, as opposed to how often or how long you've played. This distinction is only important in terms of seeding the field for high level tournaments. As most of us don't ever play in platinum events, if those high ranking points start to have some sort of shelf-life, we'll never notice.

[TimG]

hrothgar, on Aug 7 2008, 10:04 AM, said:

Some other formats (say, 12 board speedballs with 100 pairs) are complete crap shoots.

But, a series of 10 or 15 12-board speedballs should have some significance. Even though the winners of the individual speedballs will be somewhat random, the good pairs will have good results over the course of 10 or 15 speedballs, even if those good results don't include a win. And, BBO/ACBL isn't awarding any bonuses for winning a single event, the masterpoints are awarded based upon section finish.

[matmat]

rigour6, on Aug 6 2008, 01:57 PM, said:

What happens is these tournaments actually run FASTER than timed tournaments. Here's how it works, I've got 10 tables and that's about their speed. Tables 1 and 2 finish, whammo, they switch tables and move on to round 2. Everybody else is playing round 1. Then tables 3 and 4 finish, and off they zoom. Each table as it finishes, only has to wait until another table finishes and off they go to the next round. Finally only the slowest two tables remain, now this is typically more a product of disconnects than actual play but anyway, off they go.

i think i have watched tourneys like this.
I seem to recall that if two tables are significantly faster than the others, pairs might end up playing one another more than once.

[rigour6]

matmat, on Aug 7 2008, 11:36 AM, said:

i think i have watched tourneys like this.
I seem to recall that if two tables are significantly faster than the others, pairs might end up playing one another more than once.

That "shouldn't" happen, as the software should prevent it. But it's one reason why I limit these tourneys to 3 rounds. On the 2 fastest tables:
1. A v B C V D
2. A V C B v D
3. A v D B v C

No duplication. Add a 4th round, and the two fastest tables must wait for at least 2 more before you can run another round without replaying the same pair.

You could of course fine tune the software to wait until for example, 3 tables finish Round 1 and then send them all on together, which would avoid this problem. But such nuances are beyond the ken of mere TDs like meself.