[AAr]

Three Questions:

1. When should I risk going down 1 in a cold game contract trying for overtricks in IMP Pairs?

2. If I go down 1 in a cold game trying for an overtrick, about what percent of the time will such a play cause me to drop at least one place in IMP Pairs (Assume the play costs you 12 IMPs)?

3. When should I play safely for down 1 instead of risking going down 2 trying to make a game contract in IMP Pairs?

Thanks!

[Flame]

1. lets assume you are not vul, you have 420 and want to get 450, but you go down and get -50 this mean you lost 470 trying to gain 30, you lost 10 imps trying to gain 1 imp so the gain lost ratio is 1:10, therefore you should do it if the chance to succeded is greater then 10:1. if you were vul it would have been 13:1.
2. it depend on the kind of turney , how many paris, there levels, and most important how long turney.
3. lets assume non vul again, you have -50, you want to get +420, and you risk getting -100, so your gain potential is 470, and risk potential is 50, and in imps you risk 2 imps to get 10 imps, ratio of 1:5, therefore if the chances of you to succeed are etter then 1 to 5 you should try it.
vul this would be risking 100 to getb 720, therefore try it if you have more then 1 to 7.2 chance of success.

[Chamaco]

At IMPS, just make your contract, and do not worry about overtricks.

Usually try to make, even by risking a line where you would go down more.
There are some exceptions of course (e.g. they doubled you, in which case going down by a lot more will be too costly), but in "normal situations" (undoubled contracts, especially game contracts), do not fear going 2-3 down if in doing so you give youself a chance to make a difficult contract.

[Fluffy]

Not sure if this is just a fallacy, but I have teh feeling that 1 IMP normally tends into no VP difference while a real swing always does

[Blofeld]

Probably better to view the single IMP as only having a chance of producing a difference in the VPs, Fluffy ... most of the time it won't, but occasionally you'll be very glad of it. And so use the ratios that Flame suggests (I don't actually advocate trying to work out the precise chances of overtricks at the table, but it's worth knowing these figures as guidelines to aim for).

[cherdano]

Just pointing out that the odds given by Flame assume that the other table/the field will be in the same contract. If you think the field may not have bid game, there is a much bigger case for playing to ensure the contract.

Arend

[pclayton]

In long IMP matches, its usually best to go for the safety. But grinding out the extra trick can have a detrimental effect on the opponents sometimes.

In short matches scored by VP's, going for the extra trick(or 2) makes more sense than you think, especially if its a normal contract.

In IMP pairs, if the game is normal, go for the extra tricks, if it doesn't greatly imperil your contract. However, we the 'field protection' we get on online bridge this is probably dubious advice LOL.

[barmar]

Aar asked about IMP Pairs, but many of the responses here have been about IMP Team games. For the most part the strategy should generally be the same, but there are some differences. Anything you do is multiplied by the number of pairs you're comparing with, it's not just one other table.

Can someone comment on the difference in strategy between the two types of IMP events?

[SoTired]

IMHO, IMP pairs you are a little more conservative when bidding slams and generally don't bid the grand unless you can count 13 tricks. Also, the vul game bonus is not so exagerated in IMP pairs as it is in teams.

You still fight for the part score.

[barmar]

I think you should also be more conservative about sacrificing.

In teams, if you go down 500 vs the opponents' 420, you only lose 2 IMPs, so it's not a disaster. But if you're the only one who does this in a pair game, you lose 2 IMPs multiplied by the number of tables.

In general, if the field is evenly split between two results, it's similar to a team game. But any time your result is exceptional, you'll either win or lose big as a result. And unlike matchpoint pairs, a single board can have a significant impact on your final result, especially in short tourneys (like the typical 10-12 board tourneys on BBO). This is one reason I continue to play in OKbridge tourneys on a regular basis -- they have 12-board tourneys that are paired as "combos", and all the pairs that play in both parts of the combo get ranked for their 24-board results, which I think are more meaningful.

[inquiry]

I am not so sure you have the theory down just correct.... The reason why you want to be more conservative with sac's is not -500 versus -420 cost you 2 imps, it is that -300 or -500 verus +50 cost you 8 or 11, while -300 or -500 versus -420 either wins you 3 or cost you 2. The math of sacraficing at imps is quite different from that at MP. So keep in mind it is not the +3 versus -2 if you go for 2 or 3 tricks, it is also the -8 or -11 if they can't make their game.

A second place I think you have a conceptual problem is the 2 imp cost in a team game and the 2 imp cost in a pairs game. If you play in a team event, such a bid cost you 2 imps. Simple. If you play in pairs (for simplicity 10 "other tables"), and all 10 bid the game and you are the only one to sacrafice, it cost you 2 imps to each table. True this total "20", but it takes the same overtrick on two hands (winning 1 imp each time) to win back all 20 or the 2 in a team game. That is, the multiplier is neutral.... as the scoring is the same scale.

Now why you don't want to take the sac at pairs games is you can't be too certain theat game will be bid at all tables. So if you sac, rather down two or three you will lose lots of imps to those that stay out of game. And if the game happens to go down, this will amplify your lose.

[SoTired]

right, barmar, I forgot that. You almost never sacrifice in IMP pairs. A good rule of thumb, is Don't sac unless you think you have a shot at making it

[cherdano]

There is no difference between IMP pairs strategy and team strategy at all, unless you think the probability of the contracts around the field is different from the probabilities at the other table.
In a big field, there is always some table that is not in game -- but then, assuming the field is as good as the other table in the hypothetical team match, there is always a chance that the other table doesn't bid game. At IMP pairs, if you go for -500 instead of letting them play -420 you lose 2 IMPs to every table -- but you also gain 3 to every other table if you go only for -300.

So the only real world difference is that the skill level of the other tables is probably mixed in an IMP pairs event, while you may know the skill (or maybe even the system) of the other table pretty well.

Arend

[EricK]

A few points which I don't think anyone has mentioned yet:

1. How will the result of the hand affect partner? Will he get upset enough to play worse if you knowingly go down in a cold game?

2. How will the result of the hand affect you? If you are not able to immediately forget about bad results, will going down cause you to through away more IMPS later

3. How much do you need the IMPS on this board? If it is very late on in a match and you need lots of IMPS it might be silly playing for the overtrick. Your best chance of IMPS might be that others have missed the game, so you don't want to jeopardise those IMPS you have already won on the bidding. Conversely it might be worth taking a mathematically unsound risk to make a near hopeless contract if you need lots of IMPS very soon.

4. What are your chances of winning anyway? If you are one of the better pairs, it could be silly to risk throwing away a lot of IMPS going for an overtrick even if it is mathematically favourable. If you play safe and the others go for and make an overtrick, you might expect to win back that IMP later. That would certainly be easier than winnning back a handful of IMPS later. But if you are one of the worse pairs, you might feel the need to grab every IMP you can, and hope for the best.

Eric

[MickyB]

As Cherdano says, there is little difference between the tactics in teams matches and cross-imped pairs - each board in the pairs event is calculated as though you are scoring up with many teammates. In addition to his point on predicting what will have happened at other tables, there is the issue of what you are trying to achieve - if you are trying to win a short event with many tables you will need to take risks (eg not taking insurance against a slam), if you are one of the stronger pairs and you just need to finish in the top half (eg a qualifying event) then you will play down the line and be more willing to take insurance against bad scores.

Go for overtricks more readily in NV games than in vul games. Not only do you have less to lose, but 450 against 170 is worth an IMP extra (7 IMPs), whereas 650 against 170 is still only 10.

Butler IMPed pairs, unfortunately common in junior events in England, are a different matter.