[0 carbon]

Often/always in unclocked tourneys, one slow table holds up three more tables, and the slowest pairs end up playing each other at the end of the tourney.
When a table finishes a round, why not pair up the slowest pair with the fastest pair?
This can be done by keeping track of each pair's bidding + playing time in the tourney, perhaps with an offset earned in previous tourneys and stored in their profile.
Then the maximum time for each round can be minimized.
-- tom

[ArdnaXur]

Faster players already match up to next round, how can system match the slowest pair with the fastest pair?

[pescetom]

ArdnaXur, on 2024-February-21, 01:43, said:

Faster players already match up to next round, how can system match the slowest pair with the fastest pair?

Precisely.

It might be feasible to get the slowest pairs back into the pack by setting a cutoff for delay in start of round: those who do not make the cut play a round where each pair sits on the same line against a pair of robots (and then their scores are weighted against each other).

And/or assign a round not played to the slowest two pairs of all (meaning that they will come in again at the front and with the fastest pairs, at least initially).

[mycroft]

Strangely enough, the fastest pairs also end up replaying frequently in unclocked, for some reason.

Yes, "why not pair the slow players up with the fast ones?" Because the fast ones are already in the next round (or, near the end of the event, two rounds ahead). So either we slow the fast players down to the level of the slow players to give them a fast pair to play against - surely a plus for everyone! - or we don't and we get what we get - "I play the unclocked tournament, get a playback in the last round, and then go have dinner after we finish, after which the results might be ready."

(additional hint: what do you call "having the fast players wait around for the slow players to finish to pair them up"? A clocked tournament. Seriously.)

And frankly, as someone who has spent a large part of his life trying to speed up the slow players, I can't think of anything more suitable, in an environment where it "doesn't hurt" anybody else, than consigning them to playing with the other slow pairs and getting to feel what it's like from the other side. Don't like it? Play a little faster yourself, and you'll get out of the 16rpm universe.

(or, as I said in a I/N "how to enjoy your tournament more" lecture, "everybody knows what it's like to be behind the slow pair all day. And hates it. If you don't know? You're the slow pair.")

[pescetom]

mycroft, on 2024-February-21, 09:38, said:

Strangely enough, the fastest pairs also end up replaying frequently in unclocked, for some reason.

Yes, "why not pair the slow players up with the fast ones?" Because the fast ones are already in the next round (or, near the end of the event, two rounds ahead). So either we slow the fast players down to the level of the slow players to give them a fast pair to play against - surely a plus for everyone! - or we don't and we get what we get - "I play the unclocked tournament, get a playback in the last round, and then go have dinner after we finish, after which the results might be ready."

(additional hint: what do you call "having the fast players wait around for the slow players to finish to pair them up"? A clocked tournament. Seriously.)

And frankly, as someone who has spent a large part of his life trying to speed up the slow players, I can't think of anything more suitable, in an environment where it "doesn't hurt" anybody else, than consigning them to playing with the other slow pairs and getting to feel what it's like from the other side. Don't like it? Play a little faster yourself, and you'll get out of the 16rpm universe.

You missed my suggestions or you didn't like them for some reason?

The second one would also reduce the odds of the fastest pairs replaying frequently which you rightly mentioned.

[barmar]

This is a significant amount of complexity (tracking everyone's times, and storing it for every pair) to solve what seems to be a really minor problem.

Most serious tournaments are clocked, so this problem doesn't come up.

[mycroft]

No, pescetom, was just pointing out what is, not what could be (okay, until the end).

On your thoughts, I expect fewer complaints about replays and about "slow pairs playing slow pairs" - and a lot more on "why did our board get cancelled/why did we skip 2 boards" - in a tournament explicitly set to allow players to play at their own pace. I also don't expect the "play against robots" to speed them up tremendously (especially because their habit of going to the WC/getting their drink refreshed/taking a break because they called a break, even if it was just "to catch you, specifically, up" will not only not get penalized, it won't even get mentioned).

I think my key point is that I don't actually see the issue. "Speeding up slow play in unclocked tournaments" "maximum time for a round can be minimized" - why? And why penalize the faster players by not letting them play at their pace - surely they're playing unclocked because they'd rather the odd playback to avoid the annoyance of delays in the clocked events/enjoy playing at speed, and as a reward, getting paired (again) with a slow pair they have to catch up? And why penalize the slower players, who it is assumed play unclocked because they keep getting cut off/losing boards in the clocked events?

I am willing to be shown otherwise, but this not only doesn't look like a problem to me, it actually looks like a desirable solution to problems.

There are some small side-effects, certainly. For one, I can think of a player (who is definitely a M:tG Spike) who would love unclocked - he gets to play at his (glacial) pace, and get every last % out of his hands, *and* gets on average a weaker field to play against (because unlike him, most slow players are also the weaker players). But even if he wins all the unclocked events he enters, I still get a benefit out of playing in that tournament - we seed out of having to play him, and we don't have any worries about being responsible for playing quicker so that RHO can take his traditional 50% of the clock for himself even if we do get paired up.

But for me, the only issues with unclocked tourneys are the two I mentioned - playbacks at both ends, and "finish, go get dinner, then come back to see if the results are in yet."

[pescetom]

barmar, on 2024-February-21, 18:19, said:

This is a significant amount of complexity (tracking everyone's times, and storing it for every pair) to solve what seems to be a really minor problem.

Most serious tournaments are clocked, so this problem doesn't come up.

My second suggestion of merely assigning the next round not played to the last remaining two pairs would be trivial to implement, no complexity that I can see.
If you think this is a really minor problem then I suspect you never directed an unclocked tournament, nor played one with a slow partner or with a serious interest in being beaten only by the best pair rather than by the best of the slow pairs.

Sure if I was organising serious competition I would always use clocked, but the unclocked formula is a clear favourite for most players and all directors for obvious reasons. The defects of back-end slowness and lack of fairness due to repeated encounters at both ends are glaring, but could also be easily mitigated to some extent. See also next post.

[pescetom]

mycroft, on 2024-February-22, 10:29, said:

On your thoughts, I expect fewer complaints about replays and about "slow pairs playing slow pairs" - and a lot more on "why did our board get cancelled/why did we skip 2 boards" - in a tournament explicitly set to allow players to play at their own pace.

BBO already has plenty of eccentricities and tournament formats that have little do with the Laws of bridge ("You have been eliminated from this Survivor tournament", "You have 30 seconds to play and will be replaced by a robot if slower. Enjoy your tournament!") and this is just one more. "The slowest two pairs of each round will skip the next round to catch up. Enjoy your tournament!". I wouldn't complain if we were slow, I might even be grateful that we got back to the front all the same.

mycroft, on 2024-February-22, 10:29, said:

I also don't expect the "play against robots" to speed them up tremendously (especially because their habit of going to the WC/getting their drink refreshed/taking a break because they called a break, even if it was just "to catch you, specifically, up" will not only not get penalized, it won't even get mentioned).

Agreed, thinking about it only my second suggestion really holds water. Others are probably possible.

mycroft, on 2024-February-22, 10:29, said:

And why penalize the faster players by not letting them play at their pace - surely they're playing unclocked because they'd rather the odd playback to avoid the annoyance of delays in the clocked events/enjoy playing at speed, and as a reward, getting paired (again) with a slow pair they have to catch up? And why penalize the slower players, who it is assumed play unclocked because they keep getting cut off/losing boards in the clocked events?

This does not penalize the faster players at all - they still play at their pace, but when they finish the second round they might well find the slowest of the first round rather than the second fastest of both rounds. That's going to give them a fairer score, reduce playbacks and help everybody including the Director.
Nor does it penalize the slower players, except in that the very slowest get to play a round less, with no inherent score damage.

mycroft, on 2024-February-22, 10:29, said:

But for me, the only issues with unclocked tourneys are the two I mentioned - playbacks at both ends, and "finish, go get dinner, then come back to see if the results are in yet."

There is a third and equally important issue - an unfair tournament classification. The incestuous playbacks bias the classification against the fastest pairs and towards the least worst of the slowest pairs, not just because they have appropriated more time to think but because they have selected weaker opponents.

[mycroft]

I agree that the competition is a little skewed, but no more than Swiss Pairs. "you play the pairs at your speed" rather than "at your score". And sure, if your skill is in bunny-bashing, and your goal is to win (a random free or cheap tournament? Cool!), then choosing to play slowly and "select your opponents" is a strategy. Note that I can't imagine torturing myself like that for the privilege of playing a slightly weaker field slower than bridge should be played; ranks right up there with taking an hour off to do something while my partner wins masterpoints for me with a random pickup.

But surely (replays aside), having pairs "at random" play one less round (maybe two less rounds?) than the field skews the game more?

But okay, here we are, and I'm one of the faster pairs. I play the first couple of rounds in reasonable time, maybe there's a trick 1 claimer, and get paired against a pair who:

• took the "longest time" to play boards 1 and 2;
  
• got told they're slow and will miss 3 and 4;
  
• and then get to wait a few minutes for anybody to finish 4 (us) before they can play (slowly) again(*)

They're going to be so happy to see us, aren't they?

If you're not going to do it on the first round, instead waiting until the slowest table really is a full round behind the fastest - because that's less unfair - then sure, less time they "have to wait", but still.

And either we catch them up, and we get to do it again next round, or we don't, and get dropped into mid-stream for a few rounds. But eventually we will bubble back up to the fast end, and either get a pair at our speed (great), or another "skip" pair. But hey, maybe we get lucky enough to get a full round ahead of the slowest pairs, so they'd have to skip 2 sets of boards to play us, and it's only the moderately fast pairs who have to play the annoyed, frustrated ones!

And I frankly can't believe that this will speed up the tournament any appreciable amount. Okay, maybe 10, 15 minutes because those last two tables only play 16/20 and they're no longer holding up the other 30 tables, but still.

As far as "problems for the directors" are concerned, sure, it'll speed up their day, too. But so would a clocked tournament, even a clocked tournament run at "I get a chance to refill my water every round, or deal with the fact that I've refilled my water every round" speed. Whether I'd want to play in tournaments at that speed is a question; but I don't have a partner that wants to play non-sanctioned tournaments online right now, and I have no interest in the permanent pickup pool, so don't worry about me.

(*) I'm sure I'm not telling you anything you don't already know, but by Kaplan the indignity showed by the pair I've been chasing for 5 rounds to catch up, and eventually pull a board from, if the next table isn't absolutely finished and ready for them when I do. I can't imagine it will be any better when it's 2 boards they lose, and they still have to wait.

[mycroft]

I sound very denigrating. I don't mean to. Sorry about that.

You present an interesting proposal, one that if the kinks were worked out might be acceptable and helpful. But I'm still looking for the problem we are trying to solve. If unclocked is preferred "by most players and all directors for obvious reasons" then what are we fixing? And is it bigger than whatever problems the solutions cause?

[0 carbon]

One other possible solution is to automatically adjust the last table in a round once they have reached a contract.
This could be done in clocked or unclocked tourneys to speed things up.
-- tom