[Califdude]

I try to think of interesting general questions to get the thinking of the experts here on broader concepts rather than individual hands. When I posed the question whether the concept of tactical bidding was in the bridge toolbox there was very little interest, so I guess the answer is generally, no. Today I'd like to ask whether expert bridge players have the concept of pot odds in mind. (For anyone not familiar with poker, having good pot odds refers to the situation where the ratio of the pot to be won to the bet needed to stay in, is greater than the odds of drawing a card that will essentially guarantee a win.)

The application of this principle to bridge is inherent in a quotation from a bridge authority that I once read, referring to rubber bridge and assuming you bid rationally, "If you aren't going set half the time you aren't bidding enough." Makes sense with rubber bridge scoring where the game or rubber bonus vs a part-score gives you very good pot odds.

So, do expert bridge players consider the pot odds when playing IMP's. Suppose a hand comes up in which you are very sure of making 2H, probably 3H, hence you have 140 points in hand. If you make game you will score 420 or 620 so the pot odds are attractive, 3 or 4 to 1. Do the favorable pot odds influence your decision whether to bid on? Does this question deserve a wtp answer?

thanks for replies, as always.

[Cascade]

That sounds like different jargon for the concept of aggressively bidding games at IMPs especially vulnerable.

Yep wtp?

[Gerben42]

Risk / gain analysis is a part of Bridge at many points, but it's simply not called "pot odds". Must have something to do with the fact that there isn't a "pot".

[CSGibson]

This does have some application. For example, you might let a declarer play in 3 hearts vul, and decide at some point that your side is probably making 3 spades. Now you have a choice of defensive plays, one of which will guarantee a one trick set, and another that will risk declarer making it on a certain play or lie of the cards, but will set it two tricks if right (underleading an ace to get another ruff, for instance).

Playing matchpoints, you have to decide what percentage of declarers will be in the spade contract, and make a decision as to whether it is likely that you need +200 for a good score, vs just going plus. That, I think, is something akin to "pot odds" in poker.

[Califdude]

Thanks all for good replies. Not only did you respond re pot odds = risk analysis but you answered my earlier question because risk analysis as you've explained it leads to tactical bidding.

[Fluffy]

just like you might draw in poker to a flush just for the opponent to get a higher flush, or a full house, be careful with agressive bidding because you get -800 from time to time.

[matmat]

Gerben42, on Dec 18 2008, 04:29 AM, said:

Must have something to do with the fact that there isn't a "pot".

[han]

The problem with very general questions is that the answers are often either trivial or nonsense, or both. A good bidding problem will often lead to general discussions.

[JLOL]

Califdude, on Dec 18 2008, 02:51 AM, said:

"If you aren't going set half the time you aren't bidding enough." Makes sense with rubber bridge scoring where the game or rubber bonus vs a part-score gives you very good pot odds.

LOL

[barmar]

Califdude, on Dec 18 2008, 02:51 AM, said:

Suppose a hand comes up in which you are very sure of making 2H, probably 3H, hence you have 140 points in hand. If you make game you will score 420 or 620 so the pot odds are attractive, 3 or 4 to 1. Do the favorable pot odds influence your decision whether to bid on? Does this question deserve a wtp answer?

Since duplicate bridge is scored using IMPs or Matchpoints, the odds aren't nearly so attractive.

In teams, if you bid game while your opponents only bid a part score, you'll gain 6 or 10 IMPs if the game makes, lose 5 or 6 IMPs if it goes down 1. Add in a little more downside for going down more than 1 or being doubled, and you should bid game when it's 50% non-vulnerable, 40% vulnerable.

In pairs the calculation isn't so simple, because you have to estimate what fraction of the field will bid the game.

[han]

barmar, on Dec 18 2008, 03:15 PM, said:

In pairs the calculation isn't so simple, because you have to estimate what fraction of the field will bid the game.

Actually you don't.

[awm]

Here's one I've seen come up a lot.

1♥ - X - 3♥ - Pass
Pass - X - Pass - 3♠
4♥

Opener was willing to pass out 3♥, but now that opponents have competed to the three level, he has decided to bid game. Is this irrational?

Perhaps not, because the odds have changed. Suppose opener's evaluation is that 4♥ will occasionally make but usually fail (say it's about 25% to make). He sees:

3♥ gets +140 75% and +170 25%
4♥ gets -50 75% and +420 25%

But once 3♠ has been bid, the odds might change to something like:

3♠ gets -140 50% and +50 50%
4♥ gets -50 75% and +420 25%

You can see that bidding 4♥ is much better than defending 3♠, even though declaring 3♥ was better than declaring 4♥. So in effect the opponents bidding has changed the "pot odds" and made bidding a likely-to-fail game a net win.

[Edmunte1]

There are pretty similar strategic aspects in bridge as "pot odds" in poker.

In IMP's you can refer to win/loss raport. I define the win/loss raport as

R=(IMP won x % expected to win)/(IMP lost X % expected to lose). You should take the offensive action if R>=1

For example you have 3♥ in the hand (there is a big chance of making 3♥) and you think if you should bid game or not.

In non vulnerable the win/loss raport will be:
Win=420-170=250 -> you win 6 IMP's if the game makes
Loss=-140-50=-190 -> you lose 5 IMP's if the game fails
For R to be >=1, the expected percentage for bidding the game should be bigger than
5/(5+6)=45,45% In other words if the percentage of success is 45,45% R=1, W=L
W=45,45% *6=2,73
L=54,55%*5=2,73

In vulnerable
W=620-170=450 -> you win 10 IMP's if the game makes
L=-140-100=-240 -> you lose 6 IMP's if the game fails
For R to be >=1, the expected percentage for bidding the game should be bigger than 6/(10+6)=37,5%

Im matchpoints, you should refer to the concept of par. Par means the best result that can be reached by best bidding and best play. In fact there is more than one par: there is a par for N-S, a par for E-W, an absolute par, and we can talk even about "room par". Many strategic decisions can be taken related with the par.

For more informations about this issue i recommend you the book "Winning card play" by Hugh Kelsey.

[benlessard]

Pot odds arent a poker concept, theyre just a probability ratio applicable to game of chance and when you can use probability to predict random events.