[fake_user+rv@forums.bridgebase.com]

Hi,

We have a tricky situation in coming up a good format to select 3 out of 6-10 pairs (IMPs)
to represent our state in inter-state nationals later this year.

In my view if the no of tables is less than 6 (atleast 4 scores after removing top and bottom),
any form of Butler scoring is not the best.

I can think of following options:

a) Score boards (computer dealt) for every pair against datum prescribed by panel of experts
b) Play team matches wherein each pair combines with every pair (to form a team) equally (direct or 3 way matches)
c) Enlarge field by throwing it open (have more tables) and select best 3 pairs from those eligible (6-10) by Butler scores (cross-IMPs?)

What would be best option from amongst the above? Any other thoughts?

Rgds,
RV

[paulg]

In Scotland we have the same problem for the Women's trials and we have stuck with cross-imping Butler scoring. It seems as fair, or as random, as other scoring methods.

This year I believe that they are looking at scoring against a datum. This will be created by taking boards played in major championships that are seen as an equivalent standard and taking the datum from those boards. The selectors are working with another country to do this, so that the trialists will not have seen the hands.

The main downside of the datum method is the amount of work involved, but hands are easily transferred and the main scoring programmes all support this method of scoring, so it will not delay the actual trial.

[nigel_k]

I feel strongly that teams trials are best. Let players form their own teams of four and select the third pair in consultation with the winning team. There's no need to accept all entries, especially if you have an odd number. Even if six teams enter it could well be best to only let the four best teams in.

[phil_20686]

I feel strongly that in small fields team trials are not best. The main reason for this is because small fields normally imply a wide skill range among the entrants, which mean that it is very unlikely that teams formed via personal choice are very unlikely to be the strongest possible.

The psychological bias to pick ones friends is not normally outweighed by the advantage of liking your team-mates in the case where your pairs are widely spaced in ability. If your field is large you normally have lots of strength in depth. Then there are lots of team formations that are close to optimal, and the teams are normally close to evenly matched in skill level and the (relatively) small difference of getting on with your team mates will make a much bigger difference to the rankings.

[phil_20686]

In an unrelated point, small trials using datums or cross imps or butlers need some correction if some pairs are significantly weaker than others. This is because playing weak pairs on swingy hands normally results in a much bigger premium in cross imps. I think butler imps => VP helps a bit. Even just a limit to the maximum number of positive cross imps one can win in a set helps (you do not need to limit the amount a pair can lose).

[hrothgar]

A number of folks with much more experience with these matters assert that team dynamics are crucial.

It's often more important to select a team that can function well as a group than chosing the three strongest pairs. If you believe this to be true, simply let the players form teams and duke things out in some long head-to-head matches.

Alternatively, if you don't believe in the whole "team dynamic" approach and really just want to chose your best three pairs then something like the following might make sense.

Suppose that you have a total of 12 pairs competing.

1. Run a LONG pairs event with IMP pairs scoring.
2. Use the board results to generate a series of virtual team events that encompass all of the possible combinations of pairs into teams
3. Select the team that scores best across the entire set of hands

It would be interesting to see whether the results from this exercise is the same as the three highest scoring pairs...

There are weaknesses to this approach.

Most notably, in a "real" team even players can start swinging if they are far behind in a match. I'm not sure whether this same type of behaviour would work here (An individual pair can start swinging but not a team).

[NickRW]

hrothgar, on 2011-May-27, 06:47, said:

A number of folks with much more experience with these matters assert that team dynamics are crucial.

Doesn't that depend, to a fair degree, on culture. In one or two countries you have established, recognised, sponsor driven teams. In other countries there is a culture of trials and teams being made of the "best" pairs - and sponsorship is an "if only we had one!"

[phil_20686]

hrothgar, on 2011-May-27, 06:47, said:

A number of folks with much more experience with these matters assert that team dynamics are crucial.

I do not deny that they make a difference. Obviously they will make more of a difference if your players are more even to start with. One of the problems is that when people run small trials, it is almost inevitably true that there is a wide variation of skill and a number of shippers. I don't think team dynamics matter that much unless there is an extreme level of antagonism, compared to the typical variation in skill in a 6 pair trials.

[mohitz]

I might be wrong but I think when RV said he wants to select 3 pairs from 6-10, it is to select 3 pairs for a pairs tournament and not a team. It just so happens that the number of pairs to be selected is 3

[nige1]

The rise of professionalism heralds the death-knell for pairs-trials. But I agree with phil_20686

Obviously,pairs trials reduce the advantage of season-ticket-holders who would otherwise clump together into strong teams. Pairs trials have long-term advantages: Pairs trials are more effective than teams-trials at planting seed-corn, encouraging young shoots, and harvesting ripe pairs, as they come to fruition.

At Bridge, team-rapport is secondary to pair-rapport. IMO, trial entrants should sign an undertaking to play with whatever team is selected. If that deters anti-social prima-donnas, then team-harmony may even be enhanced.

The best scoring method is simple cross-imps. It is not perfect but Butlers and Datums give rise to even more anomalies.

[nigel_k]

If you restrict entries to teams of four, not six, and normally add the best performing third pair then IMO you are more likely to have the three best pairs than if you hold a pairs event.

Obviously a very long pairs event will even out the luck but they are never long enough. A couple of flat sets of cards against the weaker pairs and you are in a hole that is very hard to dig out of. Our national and provincial organisations are addicted to pairs trials so I have played in a few and I don't believe there has been a single one where the top three would be the same if you removed the results against the weakest pair. A wide skill range does not argue in favour of a pairs trial because then the randomising factor is even greater. With four teams and a semi-final or round robin followed by a long final it is very likely the winner will be the team that played the best bridge or very close to it.

Conversely, the problem of ending up with a worse team because people pick their friends is possible but in practice is not likely. As a simple example, let's say you have four pairs, the strongest is A, then B, C, and D in order, and the difference between C and D is big enough to really matter. Unless A plays with D there is no problem because the best team will win a teams trial and add the best remaining pair so you end up with A, B, and C in the team. Usually A won't form a team with D because bridge players are competitive and want to win. Even if they do there is still no problem if the team of B and C win the trial. And if pair A are really so good that they can carry inferior team-mates in a four person team, then just select them and hold a teams trial to find the other four. If you hold a pairs trial, though, you could easily end up with a team of A, B and D. All it needs is that pair D gets a flat set of cards when playing pair A and pair C doesn't.

This is before you even consider the team chemistry issue.

[ahh]

Nigel the problem with getting entrants in a trial to sign a declaration that they will play in any team is that as an entrant you may have little control over the other entrants. Bridge skill is one thing but would you choose to play in a team with a convicted criminal or someone who had failed a drugs test or someone who had been convicted of cheating at bridge or in some other area ? I know players have to pass the test of being suitable to represent their country or state or whatever but you cannot deny that a small number of undesirables (to most decent folk ) have entered trials in the past
Jim Hay