Matchpoints vs IMPs
Following is a discussion of differences in bidding between the two forms of scoring.
Majors vs Minors vs NT
MP scoring is used for pairs play, where your score is compared against the scores of all the other pairs sitting the same direction (i.e.: playing the same cards) as you. Each deal is scored separately. If you go down 7 tricks doubled, redoubled and vulnerable, you get the same zero you would get if you were in 4S making 4 when everyone else is in 3NT making 4.
IMPs scoring is used in team play, where your score is compared only against the other team's results with the same hands. It is more important to get to the safest contract than to get to the highest valued contract. If 4S is a safer contract than 3NT, it is not worth the little difference in game points to bid 3NT.
Competing For Partials
In MPs, if you compete strongly for the contract in a partscore, you may get a top for going set for "only" -100 when the opponents could have made 110 by taking the bid.
In IMPs, the difference in IMPs between a score of -100 and -110 is practically nothing, while if you miscalulate and go set for a big number, you may not be able to recover at IMPs where it's the total score that matters while in MPs, you can recover from a -1400 bottom by getting an overtrick at 1N that nobody else is getting.
Stretching For Games
At MPs, the decision of whether or not to bid game is not affected by vulnerability. If you bid game when everyone else is in a partscore, you will get a top if game makes and a bottom if it fails, regardless of vulnerabiltiy.
At IMPs, if you bid a non-vulnerable game and score 420 when the opponents stop in 3 and make 170, you net only 250 points, or 6 IMPs. If you go set one trick undoubled, you lose 50 while the opponents in the other room are making 3 for 140, for a total loss of 190 points or -5 IMPs. A 6 IMPs gain versus a 5 IMPs loss is 55%-45%, meaning that 55% of the time, you will win by bidding close nonvulnerable games.
If you bid a vulnerable game and score 620 when the opponents only get 170, the difference is 450 points, or 10 IMPs. Going down 1 vulnerable for -100 while the opponents make 3 for 140 is a total of 240, or 6 IMPS. A 10 IMPs gain versus a 6 IMPs loss is 62%-38%, meaning that 62% of the time, you will win by bidding close vulnerable games.
In the January 2005 ACBL Bridge Bulletin, August Boehm makes the point that other factors come into play, such as: (1) if you are a better defender than declarer (or vice-versa), (2) declarer normally gets more out of the cards than do the defenders, everything else being equal, so in general, it pays to declare in IMPs.
In MPs, if you are a better declarer than the average player, than you don't need to bid game to score a top if nobody else bids game. You can bid the same partscore as everyone else and get a top with an overtrick that nobody else gets.
On the other hand, if you are better at evaluating your hands, you can get a top by bidding a game or slam which nobody else (or few other pairs) bids.
In MPs, if a contested auction, your opponents may push you a level higher than pairs playing the same cards at other tables. If you are pretty sure you cannot make any more than you have already bid, then you cannot compete any higher without going set for a zero. But if you let the opponents take the bid undoubled, you may not be able to get enough points for setting them to match or exceed the points the pairs your way are getting at other tables.
So in this situation in MPs, you have no choice but to double. Even if they make their bid, you get no bigger zero than you were going to get anyway.
In IMPs, the points you get for setting the opponents undoubled, say +100 vs +140 if you had taken the bid, are not that much less than the points you would have gotten for doubling and scoring, say, 200 vs 140. However, if you miscalculate and they make their doubled contract, you will lose a ton of IMPs, so making a close double of a part score in IMPs is not a good bet.
Slam BiddingAs mentioned before, in IMPs, safety of the contract is more important than the suit. Unlike MPs, where you would stretch to bid 6NT or to bid 6 of a major rather than 6 of a minor, in IMPs, you should be guided only by which strain is the safest, even if it means bidding 6C rather than 6NT.
In MPs, if a grand slam is anything short of iron clad, you should not bid it, because there will almost always be some pairs who don't even bid the small slam, thus assuring you of an average plus in the sure small slam.
In one recent stratified club MP tournament with 13 tables, in one board EW had a 6-3 club suit, all the aces, all the kings, and the queen of trumps (Clubs) and a couple of side queens. Only two pairs bid the grand which was a laydown. One pair bid 7C and the other, 7N. The MP difference for the two contracts was 1 point. Half of the pairs in the tournament did not bid any slam at all despite having virtually honor card in the deck and a running Club suit.