[eagles123]

One thing I struggle with I think is tactics, i.e. when to bid game, when to bid slam etc. I know in imps that you should bid game even if it's fairly unlikely (is 40% right?). just wondering what the case is with MP - is there a percentage (chance of it making) at which you should bid game or slam or does it entirely depend on the strength of the field?


Thanks,

Eagles

[mycroft]

That's a very interesting topic, and entire books have been written about it...

The big thing to notice with MPs is that it's not the *scale* of the win that matters at all; it's the frequency. It's not the scale of the losses that count; it's the frequency.

At Matchpoints, bidding 3NT instead of 4M wins whenever you make the same number of tricks; so *that* has only to be a 50+% proposition. At IMPs, it only wins when the number of tricks you can take in both contracts is <10 (and only spectacularly when it is exactly 9). So you have to be much more sure of yourself.

At Matchpoints, overcalling a strong NT (there's an issue with missing game against a weak NT if you're too aggressive that counteracts these wins; the chance of game when they open 15-17 is non-zero, but negligible) on shape and "anything" is a valid tactic. It doesn't matter when it's wrong and they catch you, whether you're going 200 or 1100; it only matters how often that happens. If you convert -120 to -100, or -600 to -500, or even -120 to -110, more often than you go for your life, and if you effectively break even on tricks awarded due to their side knowing more about your hand (and it does go both ways; sometimes they fit you more likely for a card because you have more cards, and you don't have it, and you gain a trick), then you win. At IMPs, your teammates won't appreciate how often you gain an IMP or two if they have to eat more than one 800-into-partscores.

Yes, you do have to know the field:
- if you're in the field contract, scramble for that extra trick if it's +EV (and that basically means "wins more often than it loses, breakevens not counting".)
- if you're not in the field contract (good *or* bad), play safe. If you're good compared to the field, and you make it, you win already; overtricks are irrelevant. If you're bad, the only way you get anything is to hope things are bad and the field contract fails.
- an exception to the above is if you know the field contract is going to come home, you have to play to beat that score, any way you can. I get that a lot playing a weak NT in North America; as soon as dummy comes down I realize that I'm playing NT when everyone else is in hearts, and their 8 tricks are coming. +90 is going to score almost as well as -100; so I'm playing for 8 tricks even if it means I don't make 7. Similarly, I'm in hearts when they're in NT, and I know 7 tricks are easy; now *I* have to ensure it's one more in the trump fit.
- what's worse, and incredibly hard, is that you have to do this on defence as well. You have to look at the auction, dummy, and figure out what contract you're defending. Sometimes it's 5NT, even though they only bid 3. Sometimes it's "gotta beat 4♠", again even though they bid 3NT. Sometimes you have to beat 3♠ enough to make up for the +110 you had in 3♣; and sometimes you just have to go plus (because the field is going to take the push just like your opponents did, so +110 isn't going to happen very often).

I know I've mostly talked about the play not the auction, but it applies as well. Frequently you can tell you're ahead of the game in the auction; don't give it back (I remember still a hand played with a local expert, where I decided to look for 6♣. It was right to do it! ... and then I gave it all back by looking for the magic NT. Those extra 70 points weren't going to mean one lousy matchpoint, Mycroft, but going minus was 17!). Sometimes you can't tell that; sometimes you know the field isn't going to take the push, and instead of playing with them, you're thinking about gambling in the auction. If it's a good enough gamble, go for it! If it's not, play with the field and try to win MPs in the play (otherwise known as "4♠+2 on the double squeeze is worth almost as much as 6♠=").

I'll leave it to others to explain the freaky thing known as the "matchpoint double". It's *dangerous*, even when it makes sense.

[WellSpyder]

mycroft, on 2013-July-22, 18:06, said:

At Matchpoints, bidding 3NT instead of 4M wins whenever you make the same number of tricks; so *that* has only to be a 50+% proposition. At IMPs, it only wins when the number of tricks you can take in both contracts is <10 (and only spectacularly when it is exactly 9). So you have to be much more sure of yourself.

There's another way of looking at this. At matchpoints, bidding 4M rather than 3N wins whenever you make an extra trick. At imps, bidding 4M rather than 3N only matters when you can make 2 more tricks (specifically, 10 vs 8). So shouldn't playing in 3N rather than 4M be more common at imps rather than less??

[ArtK78]

WellSpyder, on 2013-July-23, 10:20, said:

There's another way of looking at this. At matchpoints, bidding 4M rather than 3N wins whenever you make an extra trick. At imps, bidding 4M rather than 3N only matters when you can make 2 more tricks (specifically, 10 vs 8). So shouldn't playing in 3N rather than 4M be more common at imps rather than less??

No, because it is usually the case that 4M will make at least one, and often more than one, trick more than NT whenever there is a good 8 card major suit fit.

[Stephen Tu]

ArtK78, on 2013-July-23, 10:26, said:

No, because it is usually the case that 4M will make at least one, and often more than one, trick more than NT whenever there is a good 8 card major suit fit.

Huh??? Yes 4M usually makes an extra trick. But WellSpyder's point is that *2 extra tricks* is considerably more rare than one extra trick. When both 3nt and 4M make 9 tricks 3nt gets a game swing. When 4M makes one extra trick, 10 vs. 9, you only gain 1 imp. Only when 4M makes when 3nt doesn't, 2 or more tricks, does being in 4M get the game swing.

So one is more likely to choose 3nt over an 8 cd major fit at IMPs that at MPs. I see a lot more game swings when one team manages to try 3nt when a bad split dooms 4M, then the other way around. Most systems are at least OK at diagnosing poorly stopped suits, or if unsure about a suit people tend to reflexively choose the major anyway.

[helene_t]

the short answer is that you should bid game when it has more than 50% chance. and the same with slam.
but if you think that the field won't be able to make game - that the >50% chance is for part due to weak opps at this table, or your.unusual system rightsiding the contract - then don't bid it. 3h+1 is already a top if nobody else makes ten tricks