[jallerton]

Suppose that:

A, B and C are all decent players of similar standard
A&B are regular partners with a decent level of partnership discussion.
C has never played with A or B before.

In a 24-board teams match in which A&B had been due to play, B has to withdraw, so C comes in to play with A. A&C have some time (say 30 minutes) to discuss system but no time to practise.

Scoring up with the same teammates, approximately how much worse, on average, would you expect the A/C partnership's IMP score to be compared with if the regular A/B partnership had played? Does it make much of a difference if A and C are aware of each other's style (e.g. if they have played against each other a fair amount)?

[Lord Molyb]

7.

[GreenMan]

I have no idea about the rest, but:

jallerton, on 2013-May-01, 15:18, said:

Does it make much of a difference if A and C are aware of each other's style (e.g. if they have played against each other a fair amount)?

I think it makes a significant difference; the more you know about what partner is likely to have in any given situation, the better you'll do overall, and understanding each other's style is a big part of that.

[ArtK78]

Lord Molyb, on 2013-May-01, 15:23, said:

7.

I don't know where you just pull a number out of thin air in answer to the question posted by OP.

I am sure everyone would agree that the answer is closer to 6.5.

[jogs]

GreenMan, on 2013-May-01, 15:26, said:

I have no idea................

That's the correct answer. Not enough info.

[awm]

It depends on how similar their general styles are. Typically if you have two expert players from the same region, they will lose quite a bit less than the other cases. The reason is that they are likely to have a similar idea of what "standard" methods are and will be roughly on the same page in a lot of auctions. Less experienced players tend to have more issues because they don't know the set of things the other doesn't know. Experts from very different places may have different ideas about some auctions that they didn't have time to iron out through discussion.

Anyway my guess would be that it's a difference of about one game or slam swing a session for expert players from the same region (so maybe 10 imps) and perhaps twice that in the other cases.

[gnasher]

I think about half an IMP per board, so roughly what awm said.

[cherdano]

I would expect no more than 0.2 IMPs per board, if A & C frequent the same circles.

[JLOGIC]

It depends on who the players are also. Some people function much better in new partnerships, and some function much better in a well established partnership with agreements. This is not a knock against either style of player but it definitely impacts the answer to your question.

[Fluffy]

in 24 boards I expect that there will be around 1 board in wich deep bidding agreements will get into play (I mean things like direct cuebid vs lebensohl+cuebid or something like)

From this 1 board, the result will matter about 70% of the time, 30% of the rest it will all be the same (All contracts down or all making, or even the wrong one is better)
From that 70% of a board they will get it right anyway 50% of the time without the agreement.

So it will matter about 35% of a board, say for an average 8 IMPs, for a total 2.8 IMP/24 boards.

Card play will be different.

[mikeh]

New partnerships between top-ranked players have an impressive win rate at major events, at least in mps.

When playing with a new partner one tends to keep things simple. And on most hands, keeping it simple can be very effective: one goes out of one's way to avoid ambiguity, where one might take more chances with an established partner and get into trouble.

In addition, the reality is that most of the more complex agreements one has in an established partnership come up very rarely and many of them afford only an occasional gain.

As one example, a partnership might play transfer walsh. The odds that a hand will arise on which transfer walsh is responsible for a pickup is low, and there will be times when it will cost by 'wrong-siding' the major suit contract, as one example. So having to revert to 'standard walsh' is unlikely to impact the score in any one round even if one accepts that overall transfer walsh affords an edge.

I'd think 30 mins is enough for two decent players, from the same area or with a common level of understanding of the game, to cover enough that in one 26 board session there will a good chance that there won't be a single hand on which a more detailed discussion would lead to a different result. I think most of us tend to over-estimate the frequency of gain from our preferred methods (a sort of confirmation bias sets in where we remember when our pet treatments worked, and these have greater significance for us than the boring majority where good fundamentals were enough).

Obviously, in any given session the practiced partnership MIGHT get an important pickup, and over the course of, say, 10 sessions, I'd definitely expect the practiced partnership to outplay and outscore 10 one-session 'new' partnerships of equivalent players.

But on average, per session, with good players I'd put the likely differential quite low: maybe 3-4 imps per session, and it wouldn't surprise me to find the new partnership outperforming the old on about 1 or 2 sessions out of the 10.

However, as Justin says, it depends on the players. In myexperience, some players are nervous in new partnerships, and those ones would probably do significantly worse, but that's because they'd be playing scared, not because of anything inherent in the agreements.

[barmar]

Sometimes players fit together well, sometimes they don't. When a pair is put together randomly, there's no telling which way it will go. On the other hand, if a pair is a long-time partnership, they presumably have a good fit, or they would have given up on the partnership. So while first-timer sometimes have good results, the chances of it are significantly less than established partnerships.

But this assumes "all else being equal". A first time pairing of champion players will still likely be better than a B-level players who have played together for 30 years. Meckwell may be one of the best partnerships in the world, but Ziawell or Meckenberg would still be formidable.

[nige1]

It depends on who A, B, and C are. We all know

• A-B long-term expert partnerships who have no detailed understandings, whatsoever. You can easily confirm that when you ask about their calls and they answer "No agreement" or "Just Bridge"
• A-C first-time partnerships, who agree "Advanced BBO Standard" or whatever and hence instantly share detailed system-agreements.

Nevertheless, IMO, an A-B partnership have a significant edge, if they are willing

• to hammer out details of effective methods that suit their personalities,
• to practice that system,
• to endure long post-mortems and, generally,
• to work at their game.

When kibitzing a Reese all stars v Sharples Gold-Cup match, I joked to Harold Franklin that it seemed one-sided. He agreed but not in the way I expected. He correctly predicted an easy win for the Sharples team because of their disciplined well-practiced partnerships and partnership-rapport.

It saves worry, if, rather than asking yourself "what can partner mean by that peculiar call?", you can, instead, ask yourself "If I made that call, what would I mean by it?"

[PhilKing]

Under the assuption that both A & B and C & A are highly compatible (otherwise the thread is pointless) I would go with a 0.2 to 0.35 imps per board figure as being the max. I would expect it to be nearer the top end of the range the stronger the opponents, since they will test your partnership more thoroughly in many areas.

edit: I have just upped my range after a rethink. Re-edit - calculations are for a very high level of understanding.

[rogerclee]

I think 0.2-0.4 imps/board sounds about right, but it depends a lot on who A, B, and C are.

[f0rdy]

My experience of playing a lot of near-pickup (a couple of emails exchanged before the game is usually about the extent of the preparation) in local leagues etc is that having fewer agreements and/or less common experience will, compared to an experienced partnership of similar standard:

• Reduce accuracy of bidding somewhat (at an IMP cost I don't feel I can guess very accurately)
  
• Massively increase variance of IMPs won or lost due to bidding. Possible mechanisms include: playing a poor but luckily making contract; being unsure about meanings of bids so sticking a game contract on the table, leaving an uninformative auction for defenders.

I'm not sure I expect the cardplay to be particularly affected¹ by a being a scratch partnership, other than in partnerships of quite unequal skill in which the better player has got used to the likely mistakes by the weaker.

¹ Apart from when (having played standard count, attitude either way for the entirety of my bridge career) I agreed to play upside down count, and even after a few sessions seemed to be using 90% of my mental effort to remember to play the "wrong" card when following suit.

[GreenMan]

f0rdy, on 2013-May-03, 05:09, said:

¹ Apart from when (having played standard count, attitude either way for the entirety of my bridge career) I agreed to play upside down count, and even after a few sessions seemed to be using 90% of my mental effort to remember to play the "wrong" card when following suit.

Interesting; a while back when my partner and I, neither of us much experienced at the time, decided to switch to UDCA, our defense immediately improved, not because the new signals were that much more efficient, but because we were focusing a lot better because of paying attention to each card. I suspect switching count but not attitude, as you suggest, would be more problematic.

[barmar]

GreenMan, on 2013-May-03, 10:52, said:

Interesting; a while back when my partner and I, neither of us much experienced at the time, decided to switch to UDCA, our defense immediately improved, not because the new signals were that much more efficient, but because we were focusing a lot better because of paying attention to each card. I suspect switching count but not attitude, as you suggest, would be more problematic.

Studies have shown that you have a limited amount of conscious mental energy. If you have to use more of it to remember how to send or read a message, it's likely to be at the expense of deciding what message you want to send or what the received message means. Also, you're less likely to make a mistake when performing an automatic action than one that requires conscious thought.

Those are generalizations, in some cases the conscious thought may reinforce things, as you describe. But in general, your play gets better as more of the routine actions become automatic, and that requires them to be familiar.

[the hog]

ArtK78, on 2013-May-01, 16:01, said:

I don't know where you just pull a number out of thin air in answer to the question posted by OP.

I am sure everyone would agree that the answer is closer to 6.5.

Disagree strongly - 3.2

[Antrax]

It just gets funnier the more you spam the topic with these replies.