Matchpoint vs IMP

[dcvetkov]

I read some time ago somewhere interview with Fred Gitelman in which he mentions that he likes matchpoints better, because its a better test of bridge skills of one player. I like that each boards has equal importance and you can still recover from disaster, which is not true at IMP. There are also things to consider at card play and defense, where objective may be different from IMP

 

Any comments on this? It is generally known that amount of random luck is greater in MP then IMP . ie team of 4 will beat weaker team large percentage of the time over sufficient number of boards). ( There was a precise statistic about this, but I dont remember it).

 

Which on do u prefer? What about Pairs IMP scoring ( Cavendish may have unique scoring, but I assume it belongs here and there were comments from Vugraph commentators that lot of people dont like the format since there are huge swings and takes quite a bit of luck to finish high).

 

Ok Ill throw here rubber bridge as well, just for fun;)

[Echognome]

For me IMPs. I enjoy team play more than pairs play. I agree they are different skills used. I just enjoy figuring out the best odds to make contracts (allowing for safety) and also being able to be more aggressive on defense. To me it's more exciting.

[awm]

I like BAM the best. It's not particularly close.

 

In general IMP scoring seems pretty random to me. You can make a great defensive play but the contract is just cold and you get 1 IMP, or you can miss a dicey slam that makes and lose 13 IMPs. The fact that boards don't count the same means that the match often comes down to a small number of boards where different reasonable decisions were made, rather than actual skill. Obviously playing a very long match will reduce this effect.

 

Of course, MP pairs scoring in a weak field is pretty frustrating too, because it depends so much on getting gifts and/or taking advantage of the bad players. MP pairs in a strong field is much more fun than IMPs for me, but strong fields are few and far between.

 

BAM scoring is really best of both. You have teammates to protect you, so you're not subject to the swings of "the field" but at the same time every board counts equally and the match won't come down to one or two random decisions that happen to have higher value than all the other boards.

[Trumpace]

BAM is good, but shouldn't the penalty of not bidding a makeable grand slam be higher than not making a mere overtrick?

 

Perhaps if we scaled down the IMP scale non-linearly, depending on the number of boards played...

[Fluffy]

I play MP most of the time, it helps on the way that every mistake count so you train your concentration, it can be unfair sometimes.

 

Don't like the big loss for +50 against partscore, -200 agaisnt partscore and -500 against game.

 

 

I don't like from IMPs that 50% slams decide that much, or that finding the trump queen on a doubled vulnerable contract might cost 24 IMPs.

 

 

 

I played in France a couple of times a scoring format that combined BAM with IMPs, it was funny.

[P_Marlowe]

BAM is good, but shouldn't the penalty of not bidding a makeable grand slam be higher than not making a mere overtrick?

 

Perhaps if we scaled down the IMP scale non-linearly, depending on the number of boards played...

Or try Patton.

 

With kind regards

Marlowe

[MickyB]

Over a specific number of boards, MPs is less random than IMPs, with total points being more random still. This is due to the relative values of boards - your result matters on every board at MPs, whereas at IMPs overtricks are relatively meaningless. At total points, even a NV game swing is worth very little compared to the grand-slam swing on the next board!

 

Teams vs Pairs is interesting. A much better team is very likely to win a teams match partly because, even if one of their pairs produces below average results, the chances are that their teammates will make it up at the other table. I feel strongly that national trials should be pairs-based - why should your pair represent your country just because your teammates did well?

[Fluffy]

why should your pair represent your country just because your teammates did well?

I've heard this before, but you should take into account that a pair will do better teamed up with one they like than with one they hate each other. You have to live with them for a couple of weeks or so.

 

There is a Big difference in there, more than you may think for first.

[Guest Jlall]

I used to love MP but now....not so much :)

[helene_t]

MP (or BAM) is statistically more stable since all boards count equal. If you play 23 dull 1NT or 3NT boards that may or may not produce an overtrick depending on skills and luck, and one vulnerable grand that may or may not make depending on skills and luck, any luck factor deciding the grand is likely to decide the match at IMPs. At MPs, any difference in skills between the two teams is likely to decide the match due to the law of large numbers.

 

The idea that more luck is involved at MPs than at IMPs is strange. Either something is wrong with my reasoning, or some psychological mechanism is at play (if you make the grand on an anti-percentage line you may attribute it to skill rather than luck, while if you pass out a board where the field bid 3N down one you may attribute it to luck), or it's just another urban legend.

 

BAM can be frustrating, though, when for example you put a lot of effort into an overtrick only to learn that the contract wasn't even bid at the other table. This frustration is irrational but I can understand it, in fact I get frustrated by BAM as well.

 

MP can be frustrating when you play in a small, heterogeneous field. You have no clue if you must play safe or not if you don't know whether the Godzilla's or the experts are in the same line as you. In principle you could study the list of participants and factor the line-up (whether known or considered a random variable) into your calculations, but that's unrealistic for most of us.

 

I prefer IMPs, for two reasons. First, my awareness is not good enough to concentrate on overtricks all the time, especially when defending. Second, I have troubles enough with the basics so I don't need the extra challenge of assessing what the field is doing. Sabine Auken wrote in "I love this game" that some matchpoint freaks can estimate the board statistics with astonishing accuracy (at least in a large, strong, homogeneous field) but that she can't do it all.

 

Now if Sabine Auken can't do it I'm not afraid of admitting that I can't do it either. But in theory, MP should be fairer and more interesting.