Once in a lifetime?

[JLOL]

I was looking over the hand records from a club game and it had deep finesse analysis of how many tricks each side can make in each denomination fromm each declarer. If you cannot make 7+ tricks in a denomination it just leaves i blank. Anyways there was a hand where no declarer can make anything...literally. Par on the hand is passout.

 

Anyways I was just wondering if I was right in how unlikely this is, it seemed super unlikely like a once in a lifetime thing. Am I overestimating the improbability of this? Can anyone smart come up with an estimate of how likely this is or is that too difficult?

[helene_t]

I didn't find any par zero deals among the first 10,000 deals in the GIB DD database. Gerben once posted a link to a website that gave some examples.

[jdonn]

I once saw a hand on the hand record where on double dummy play and defense, NS make 9 tricks in all five strains. Somehow that seems to me even less likely, but who the heck knows.

 

BTW can you post that hand? There was a thread like 2 weeks ago where someone asked if that very occurence was possible.

[TimG]

I didn't find any par zero deals among the first 10,000 deals in the GIB DD database. Gerben once posted a link to a website that gave some examples.

That's not even a year's worth of bridge for Justin.

[Echognome]

Try here for some examples:

 

http://bridge.thomasoandrews.com/deals/parzero/

[vuroth]

First one on that site depends on the lead. Granted, the lead seems obvious.....

[kenrexford]

I have seen many deals at club games where no one seems to make anything. Of course, that might be based on less than deep finesse declaring.

[nickf]

1) I am sure Richard Pavlicek has constructed many such deals.

 

2) Having the little DF box blank doesnt always mean no contract is makeable. On old machines, and with complex hands (usually flat part scores where the HCP are evenly distributed), the DMPro analyser can take literally hours to find the makeable contracts. There's a keyboard option for the operator to skip the analysis on that hand, thereby leaving the analysis blank.

 

nickf

sydney

[thomaso]

First one on that site depends on the lead. Granted, the lead seems obvious.....

 

It's not clear what you mean by that. Nobody is asking for a deal where no contract makes on any lead. We are talking about double dummy results.

 

--

Thomas

[Fluffy]

Pavlicek had a lot of research on this, he looked for years for a hand where everyone made 1 of a suit, which is much harder, every possible lead had to give away a trick double dummy. He finally found one.

 

Funniest for me still is the one where every player makes 3NT.

[thomaso]

Pavlicek had a lot of research on this, he looked for years for a hand where everyone made 1 of a suit, which is much harder, every possible lead had to give away a trick double dummy. He finally found one.

 

Funniest for me still is the one where every player makes 3NT.

 

To be more accurate, he looked for a hand where all seats can make 1 of the same suit.

 

And here is the deal I found which satisfies this condition. In this deal, every seat makes 1N, 1S, and 1H.

[Zelandakh]

A deal where every seat can make exactly one of any denomination would be particularly special.

[Vampyr]

1) I am sure Richard Pavlicek has constructed many such deals.

 

2) Having the little DF box blank doesnt always mean no contract is makeable. On old machines, and with complex hands (usually flat part scores where the HCP are evenly distributed), the DMPro analyser can take literally hours to find the makeable contracts. There's a keyboard option for the operator to skip the analysis on that hand, thereby leaving the analysis blank.

 

nickf

sydney

 

Also the settings on DF matter. Sometimes it only looks at cases where the putative declaring side have a majority of trumps, some proportion of the HCP etc. I tend to trust only DF analyses which give a number for the tricks each player can make in each denomination.

 

With different settings I have seen boxes that are inaccurately blank.

[thomaso]

I have a personal database of 6,000,000 deals, of which 15 are par zero, which puts, in this sample, the probability at 1 in 400,000.

[HighLow21]

I have a personal database of 6,000,000 deals, of which 15 are par zero, which puts, in this sample, the probability at 1 in 400,000.

i.e., about 1.25 times in a lifetime (assuming you play half a million hands in that life of yours).

 

Of course, the confidence interval on the sample (assuming each hand is independent) is 0.60 to 1.89, so who are we to say it is not indeed once in a lifetime?

[CSGibson]

I have a personal database of 6,000,000 deals, of which 15 are par zero, which puts, in this sample, the probability at 1 in 400,000.

 

 

How are hands added to your database? Randomly generated or self-selected in some way?

[thomaso]

How are hands added to your database? Randomly generated or self-selected in some way?

 

Entirely randomly - no filter applied.

[thomaso]

The only parzero deal in my sample with a 9-card fit or better:

 

[hv=pc=n&s=st72hjdakt872ct93&w=s9653h8642dqjcqj2&n=sk84hakqtd543c864&e=saqjh9753d96cak75]399|300[/hv]

[Zelandakh]

I'll bet a coke this hand was not passed out at the table.

[Lord Molyb]

I'll bet a coke this hand was not passed out at the table.

that reminds me of a hand in which the par result was E/W playing 4♠X-1 in their seven card fit when they have a side 9 card fit and a combined 16ish points

[thomaso]

that reminds me of a hand in which the par result was E/W playing 4♠X-1 in their seven card fit when they have a side 9 card fit and a combined 16ish points

 

Bah, here's a deal where double dummy par is 6♦X-3 with a 6-card fit and a 7-0 split.

 

Part of my Bad Fit Deals collection, where par is to play in a six-card fit or worse.