[Chamaco]

Hi all,
last week I had a bridge conversation about a hand with a real expert.

He is marked as star in BBO, has consistently participated to international events, writes a bridge column etc etc, so, shortly, he cannot be dismissed as one of those "BBO selfrated experts", who most often would be better labeled as "experienced".

We were talking about a game contract, where there was available a safety play vs the 5-0 trump break.

This safety play involved surrendering an almost sure overtrick, ti guarante 100% 10 tricks in a major suit contract.

He said that- even at IMPS - the cost/benefit of these safety plays does not justify them, the reasoning being: say that a game is worthwhile 11 imps or so, and an overtrick 1 imp, these safety plays would be justified only if the risk they want to avoid is equal or higher to more or less 1/11.
1/11 = about 9%, and if we want to be conservative, we can reduce the % to 5% , meaning that a safety play will be chosen if it avoids a 5% risk, but not to avoid a risk of, say, 2-3%.

He says that al these 1%, 2% , 3% losses of overtricks tend to add up in the long run (EDIT: not in the single match, but over many matches), more than a single bad break that causes an 11 imp loss.

I replied that, according to some textbooks, when the bad break indeed occurs, it causes a loss of more of 11 imps, in terms of confidence, and harmony with pard and teammmates (EDIT: and does the opposite to opps, boosting their selfconfidence).

He says that this is a feeble argument and that between two good players this should not occur.


Comments ?

[Trumpace]

Say the chances of going down are 5% if we are trying for the overtrick. Suppose a game swing is worth 11 IMPs. Assume that at the other table they always play safe.

if you take the risk and it works you end up making +1 IMPS.
if you take the risk and it loses, you end up making -11IMPs.

Your expected IMP score per board is 0.95*1 + 0.05*(-11) = 0.4 which is positive, so you must risk the overtrick. Of course this assumes your opponents always play safe.

In fact, if the chance of going down when trying for an overtrick is around 9% (100/12 to be precise) or less, the risk must be taken.

But the problem with this analysis is the usage of the words "in the long run". The frequency of occurence of such boards in a say 256 board match could be really low. For each board you throw 11IMPs, you need 11 such boards to catch up, what are the chances of that? That needs to be taken into consideration too!

In the long run counts only if you are playing a reasonable number of boards which decide the outcome of a match! which could well be over a 1000...

This post has been edited by Trumpace: 2005-December-19, 03:34

[awm]

It often depends on state-of-the-match type considerations. If you believe your team is substantially stronger, or that you have a substantial lead already, then taking safety plays is right. Losing an imp here or there will not decide the match, so you avoid the possibility of a large adverse swing. Also, it may depend on how likely you think the other table is to reach the same contract -- if your in a pushy game that they are likely not to bid, then best to make it.

But against a roughly equal team, in a close match and a normal contract, it's better to play for the overtrick when the safety play is guarding against a very unlikely position.

I agree that the "psychological factors" mentioned are a very weak argument.

[Chamaco]

awm, on Dec 19 2005, 09:36 AM, said:

Also, it may depend on how likely you think the other table is to reach the same contract -- if your in a pushy game that they are likely not to bid, then best to make it.

Yes, the discussion was about:
- "normal obvious" contracts, e.g. we might be pretty sure it was bid at the other table.
- State of the match, normal. In short, nothing unusual would be called for at this point in the match, just optimal technique.

[Trumpace]

Going further on the number of 'Overtrick or die' (OOD) boards in a particular match.

Say a 100 board match has 11 OOD boards (which is probably unlikely) What should your strategy be?

Say the chances of going down on a single board are 5%.

The chances that you gain an IMP on all the 11 boards is (0.95)^11 = 0.568 ~ 57%.

The chances that you lose at least one IMP total (by going down at least once) on the 11 boards is ~ 43%.

If you think 43% is high enough, play safe.
If the chances of going down on a single board were 8% this figure would be a whopping 61% (= 100*(1- (0.92)^11)).


Now consider a match where there are 23 OOD boards.

The chances that you go down on at least 2 boards is: 1 - (0.95)^23 - 23*(0.95)^22*(0.05) = 0.33 ~ 33%

With 35 OOD boards the chance of losing at least one IMP drops to ~25%. (if chances for a single board were 5%)

In the long run, the chance of losing IMPs is less than 50% (as shown by earlier analysis)

But the long run, might not be long enough... as the 8% case for the 11 boards shows.

This post has been edited by Trumpace: 2005-December-19, 13:12

[Robert]

Hi everyone

Some experts have some very strange bidding ideas. Just look at bidding contests

with a top level expert pair. One says, "sign off', the other says, "invites slam."

In a long match of 64 boards, his advice would determine the play in very few

boards(looks like slightly more than 3% chance for the 5-0 break and this would

only apply to game contracts with 5 outstanding trumps, since hands with 9+

trumps would not count in our survey)

Winning one imp 1-2 times vs losing 11 tricks "once" is a gamble that I would not

take.

Empty a revolver, replace one bullet and spin the chamber. The odds are six

one that you can place the pistol against your temple and 'win.'

A single loss will "not" make up for the 'one' losing option.

Assuming the experts at the other table are following 'book' play, you will

always tie by taking the safety play. By gambling you might win one IMP, but the

penalty for being wrong is a game swing. "Death Wish" type tactics at IMPs.

Regards,
Robert

[whereagles]

The odds for a 5-0 break are around 4%. The overtrick costs 1 imp out of 10. That's 10%, so indeed in theory it doesn't pay to play for the 5-0 break.

If the match is close, play for the overtrick. If you're winning and just want to avoid big swings, play for the 5-0 break.

[Fluffy]

I have though the same view, althou never found someone who agreed on that .

At Cross IMPs (butler) overticks are better than team matches.

at Team matches with VP, you have to take into consideration that 1 IMP often leads to no VP, it costs you 0-1 VP. But 10-13 IMP mean 2-4 VP.

When you only have to win the match because its a ko round then you can think back like cross IMPs.

[Trumpace]

whereagles, on Dec 19 2005, 06:07 AM, said:

The odds for a 5-0 break are around 4%. The overtrick costs 1 imp out of 10. That's 10%, so indeed in theory it doesn't pay to play for the 5-0 break.

If the match is close, play for the overtrick. If you're winning and just want to avoid big swings, play for the 5-0 break.

If you follow this strategy... and say we play matches with 17 such board each. You will end up losing more often than winning. (Assuming that losing a 4% 5-0 break board throws away 17 IMPS, which according to your logic odds are around 5% in theory and is not worth playing safe).

The problem with this 'long run'/in theory analysis is that the IMPS won on such boards by USA against Italy in Bermuda bowl 2003 are not counted in USA vs Italy match in the Bermudal Bowl 2004. The long run analysis has an assumption of such a carry over!

[vang]

Trumpace, on Dec 19 2005, 11:14 AM, said:

But the problem with this analysis is the usage of the words "in the long run". The frequency of occurence of such boards in a say 256 board match could be really low. For each board you throw 11IMPs, you need 11 such boards to catch up, what are the chances of that? That needs to be taken into consideration too!

In the long run counts only if you are playing a reasonable number of boards which decide the outcome of a match! which could well be over a 1000...

i think Trumpspace pointed very well the problem. most of the time i play (very) short matches (since i didn't made yet in BB final ;-) and to loose 11 IMPs trying to gain 1 in one board would be crazy.

[Gerben42]

Back to psychology. For you and your partner it may or may not make a difference (it shouldn't) but there is a factor you cannot control.

If you make the safety play and it was necessary, it will be a blow to your opponents who are imagining 11 IMPs in the "them" column. They will play worse.

Otoh if you go down where a safety would have seen you home, they will imagine 11 in the "us" column, thus their spirits will go up. They will play better.

Just make your contract.

[whereagles]

Trumpace, on Dec 19 2005, 11:47 AM, said:

whereagles, on Dec 19 2005, 06:07 AM, said:

The odds for a 5-0 break are around 4%. The overtrick costs 1 imp out of 10. That's 10%, so indeed in theory it doesn't pay to play for the 5-0 break.

If the match is close, play for the overtrick. If you're winning and just want to avoid big swings, play for the 5-0 break.

If you follow this strategy... and say we play matches with 17 such board each. You will end up losing more often than winning. (Assuming that losing a 4% 5-0 break board throws away 17 IMPS, which according to your logic odds are around 5% in theory and is not worth playing safe).

Well, a 5-0 break doesn't throw away 17 imps but something like 10 or 12.

Anyway, if you always want to "play for the contract", you might one day lose an international match 20-10 VPs, the same way I once did. Me and pard kept playing safe while on the other room opps kept trying for overtricks. After 20 boards those peanut imps added and we lost by a substantial margin.

[hrothgar]

This is simply a special case of the variance / expected value issue discussed in the other thread...

[Flame]

vang, on Dec 19 2005, 06:48 AM, said:

Trumpace, on Dec 19 2005, 11:14 AM, said:

But the problem with this analysis is the usage of the words "in the long run". The frequency of occurence of such boards in a say 256 board match could be really low. For each board you throw 11IMPs, you need 11 such boards to catch up, what are the chances of that? That needs to be taken into consideration too!

In the long run counts only if you are playing a reasonable number of boards which decide the outcome of a match! which could well be over a 1000...

i think Trumpspace pointed very well the problem. most of the time i play (very) short matches (since i didn't made yet in BB final ;-) and to loose 11 IMPs trying to gain 1 in one board would be crazy.

I think if anything then its exactly the opposite, if the match is short you dont want to throw overtricks , because in the long run it will benifht, because the long run wont come, the trunps will not break 5-0 not at this match.
All in all the expert is right, take the matematical right play.

[helene_t]

This psycological argument is strange. If there's a 96% chance of an overtrick there's a 96% chance of giving the teammates the feeling of defending better than the opps at our table. Hopefully irelevant, but how could it be a bad thing?

[ArcLight]

>He said that- even at IMPS - the cost/benefit of these safety plays does not justify them, the reasoning being: say that a game is worthwhile 11 imps or so, and an overick 1 imp, these safety plays would be justified only if the risk they want to avoid is equal or higher to more or less 1/11.
1/11 = about 9%, and if we want to be conservative, we can reduce the % to 5% , meaning that a safety play will be chosen if it avoids a 5% risk, but not to avoid a risk of, say, 2-3%.


This was addressed in one of Kelsey's (or maybe was it Klingers?) books.

Obviously the state of the match is of paramount importance.
Other than that, you can reason:

my expected gain by not taking the safety play is + .X (some small number of IMPS)

However, if you go for the over trick, and go down, you may damage your partners morale, the teams morale, and harm your relationship with the team.
Most other teams will have gained many IMPS on you, this one time. That may be enough to lose this one match. While gaining an expected 0.5 IMPS is unlikely to be important unless you have the chance to make many such decisions over a large number of hands.

In short, the author (I think it was Kelsey) sugegsted taking the safety play, for the interest of long term partnership and team morale.

[Gerben42]

Because when you are comparing, opponents have already escaped from your table so their state of mind will not be changed by anything except the total score (of which your overtrick is an irrelevant part). However if you make the safety play and it was necessary, opponents are still at your table, ready to feel that they have lost 11 IMPs and ready to give you more IMPs!

[42]

Interesting topic! Do wc players and experts at the table in a long hard match against equal level players guard against a 5-0 break when the bidding was quite normal (let's say: uncontested, or non jump or 2-suiter interventions) and the contract is expected to be the same in the other room? Bridge is also a game of probabilities and it is more likely that a suit breaks 4-1 or 3-2 than 5-0. We've all seen matches who were won (or lost ) by 1 IMP, why should one assume that it will be important to make the safety play in a "normal" board when it is not the "practical" action? The board is not yet decided until the player in the other room chooses a different line for some reason, but why should he? So it may not only be a question of technics or mathematics but also of psychology. The players we talk about have shown their will and ability to win, I think they do not have to explain in detail why they didn't play for the 5-0 break when it would have been the winning line; and I do not think (or I hope...) that a single 11 IMP gain or loss has any influence on the moral of any high level team, in opposite perhaps to weaker teams or short distance matches.
I remember that when I read a book about endplays or so, I tried to use my new knowledge in nearly every board and that did me no good It is an art to do the right thing at the right time against the right opps... One question still remains: what IS right?

[mikeh]

I agree with much of what has been written. However, the sense I have got through my own experience at the table and from reading many writeups of big matches in decades of Bridge Worlds is that the majority of good players take the safety play unless their table feel tells them that the overtrick play will definitely work.

One of the reasons, I think, is that there are very, very few hands on which one can be absolutely certain that the other table is in the same contract. I have not attempted any form of statistical analysis but my guess is that one or more imps changes hands on the majority of boards even in a match against two comparable teams. Kaplan long espoused the suggestion that, for example, one bid a grand at imps only if confident that it was at least a 75% contract: a figure far in excess of the theoretical break-even point. His argument, proven true on countless hands in high-level competition, was that the opps at the other table might miss a seemingly easy small slam.

So while there is definitely solid logic behind the play for the overtrick on mathematical grounds, I think the weak point of such an approach is found in the assumption that the contract is 'normal', whatever that means.

My comments are aimed at long(er) matches: 28 boards or more. If you are playing primarily 7 or 8 board matches, I'd still take the safety play against most teams, because I expect to win at least one or two decent swings and want to avoid a disaster against which 1 imp is a modest premium. Against a tough team, in a short match, I'd risk the overtrick if my antennae were in good shape that day (they usually are not: I'm not much of a 'feel' player, alas)

[pclayton]

It certainly depends on the length of the match too. In a Swiss team, I'm less likely to safety anything, but in a KO, I might.

If you've never lost (or won) a match by exactly 1 IMP, then please don't participate in this discussion.