[aguahombre]

Cross IMPs? datum comparisons? Probably different answers for the two.

Assuming a decent quality field, how much do the IMP odds change for game decisions, sacrificing, penalty doubling, etc., when we go from teams to pairs?

In, for instance, the ACBL Natl. IMP Pairs event, wins and losses on a particular board seem to be about half what we would expect for a swing at teams.

OTOH, minus 2 IMPs on a board puts us 4 IMPs down to a pair who were +2 on it.

[Trinidad]

The way I look at it is that I am playing a -relatively long- team game against every table in the room. The quality of my team mates and opponents varies. The weaker opposition may mean that you might want to be a little less aggressive in bidding games than you would be in a team game against strong opponents. However, the fact that the match is relatively long (28 boards or so) means that you should be playing as if you play a team game against strong opponents. There will be enough swings so you should go with the odds.

Rik

[Mbodell]

I think it is basically the same. The swing is the same in both, it is just in teams you usually think of it as from 0 IMPs when you both bid and make game to -10 IMP when you are 170 into 620 for a spread of 10. In the IMP pairs if the hand was a borderline one that basically half the field is in part score and half is in game with everyone making 10 tricks then those in game will get +5 and those in part score will get -5 and it is the same range.

[barmar]

The difference in IMP pairs is that an unusual result gets multiplied by all the pairs, while common results get diluted.

For instance, suppose there's a board where everyone bids game, and success hinges on guessing a 2-way finesse. Half the field will get it right and get +5, half get it wrong and get -5. If you're in the winning half, you've taken a big lead against the losers, but you're even with the other winners.

On the other hand, if there's a board where the normal play goes down, but you got helpful defense and are the only ones to make it. Now you get +10, and everyone else gets 0. You've taken a big lead against the entire field.

You tend to notice this more when you're on the losing side of a decision. If you overbid to a no-play slam, and no one else even sniffs at it, it's a killer.

[paulg]

The Scottish trials are cross-imped pairs and, disappointingly, most of the competitors say that it is very similar to playing in the main bridge club on BBO. Unless it is a top field (like the Cavendish) you get 2 imps for bidding the most boring game so there is a fair amount of luck. The strategy is the same as teams (imps), but the big boards tend to more important.

[helene_t]

Trinidad is right.

We had the same discussion about BAM vs matchpoints pairs. The scoring is the same in XIMP as in IMP Teams just like BAM is the same scoring as MP Pairs, so any differences between teams and pairs (given that the scoring is the same) boils down to
- different objectives. You may have a different attitude to normal vs speculative actions when your objective is to win a team match, as opposed to when your objective is whatever it may be in a pairs tourney.
- guesses as to what the players at the other table might do. Obviously easier at team when you know who is sitting in which direction at the other table.

Anyway, these differences are tiny compared to the difference between MP and IMPs. The best advice is probably not to worry about it and just play normal bridge.

Butler (typically IMPs versus a truncated average, although other butler scorings exist) is slightly different from IMP/XIMP:
- If you go for 1400 on a partscore hand, you get your minus but you will be truncated from the datum so you competitors don't get the full value. So it is slightly less of a disaster at butler than at IMP/XIMP.
- If you are in an atypical contract, at IMP/XIMP you will "know" that an overtrick doesn't matter, but at butler it could matter because even if your score is atypical it could still be close to the datum. So when you are in an atypical contract, overtricks matter more at butler than at IMP/XIMP.
- If you are in a typical contract, at IMP/XIMP you expect close to one IMP. But at butler, there will often be a single maverick who drags the average somewhat away from the modal score which might destroy the impact of the overtrick. So in typical contracts, overtricks matter less at butler than at IMP/XIMP.

All this said, I think the difference is so small that it isn't worth worrying about. Total points is very different from matchpoints, and IMPs is somewhere in between, but other than that: forget about the scoring and just play bridge.

[phil_20686]

The only really important thing IMO is to remember that over tricks are really quite valuable at cross imps.

This is because the variance is lower. When you play 24 boards against one team there are normally enough 10 imp swings that the number of over/under tricks is close to irrelevant. When you play against ten teams its likely that you are close enough to some other teams that over tricks will move you up or down by one or more places. We care about places rather than total scores.

This is similar to playing in the MBC, if bidding a normal game gets you + 2 imps, an overtrick turns that into +3 imps, and they really stack up, because your typical score on a board has a much smaller variance.

[mycroft]

Having played through the "play the Cavendish with me" CDs, one thing I noted was that "going plus" was worth about +1.5 IMP/board (+.5 when "everybody is going plus" and +2 when not "everybody is going plus"). So I'd guess that at a non-Cavendish level IMP pairs, that might be even magnified a bit.

[aguahombre]

mycroft, on 2011-November-21, 12:26, said:

Having played through the "play the Cavendish with me" CDs, one thing I noted was that "going plus" was worth about +1.5 IMP/board (+.5 when "everybody is going plus" and +2 when not "everybody is going plus"). So I'd guess that at a non-Cavendish level IMP pairs, that might be even magnified a bit.

sounds like an argument for matchpoint strategy at IMP pairs, if it is an observation which holds true over the long run.

[S2000magic]

phil_20686, on 2011-November-21, 05:47, said:

The only really important thing IMO is to remember that over tricks are really quite valuable at cross imps.

This is because the variance is lower. When you play 24 boards against one team there are normally enough 10 imp swings that the number of over/under tricks is close to irrelevant. When you play against ten teams its likely that you are close enough to some other teams that over tricks will move you up or down by one or more places. We care about places rather than total scores.

This is similar to playing in the MBC, if bidding a normal game gets you + 2 imps, an overtrick turns that into +3 imps, and they really stack up, because your typical score on a board has a much smaller variance.

So, is the strategy at IMP pairs the same as at masterpoints, or maybe a hybrid between IMP teams and masterpoints?

[phil_20686]

its obviously not worth risking your contract at imp pairs,

but often when playing real teams when you have 11 on top people virtually switch off. Or more commonly when you have beaten a contract by one you have to think how to beat it by two.