[ArcLight]

When (if ever) would you risk your contract for an overtrick at IMPS?

Would you risk a game for an overtrick if the success rate was 95%?
(say a 4-0 split, with one specific player having the void, and no information about distribution - xx opposite AKQxxxx, if LHO is void, you duck, else do you try for the overtrick with no other entries?)

It seems that over 20 games you would come out ahead doing so. These types of decisions might not come up so often. And it would be a real downer to lose a match by a few IMPS because you went down in a cold contract searching for 1 IMP.

[jtfanclub]

http://forums.bridge...showtopic=20104

[Stephen Tu]

If you are sure the other team/field will reach same game/slam, then yes, it's worth the risk when the overtrick is likely & chance of booting the contract is so low, in that 5% range. These are pretty rare though. Especially for slams don't give the other team too much credit for reaching it, accidents happen all the time.

Also, one should take even bigger risks if say you were doing a 2-board tiebreaker in a KO match (no onerous safety plays!), or toward the end of a short Swiss match w/ win-loss (not ACBL 20 VP scale) scoring towards the end if you were certain there hadn't been any swings yet.

[hotShot]

Lets do some math here:

The other table makes game +420 you go down 1 -50 => you loose 470 => 10 IMPs
The other table makes game +420 you make one more 450 => you win 30 => 1 IMP
This means in non vul. both you need to make 91% of the overtricks to brake even for the loss you make, if the contract fails.

The other table makes game +620 you go down 1 -50 => you loose 670 => 12 IMPs
The other table makes game +620 you make one more 650 => you win 30 => 1 IMP
This means in vul./? you need to make 93% of the overtricks to brake even for the loss you make, if the contract fails.

[P_Marlowe]

ArcLight, on Jul 9 2007, 02:02 PM, said:

<snip>
It seems that over 20 games you would come out ahead doing so.  These types of decisions might not come up so often.  And it would be a real downer to lose a match by a few IMPS because you went down in a cold contract searching for 1 IMP.

No, there is another reason, matches are played with
a finite number of boards.
There exist an little amusing mathematical problem, which
is related to your question:
Two friends play regular tennis matches against each
other (best of 5 sets), one wins on average 2 out of 3 points,
how likely is it, that the other player wins a complete match.
A hidden remark follows.

Spoiler

I dont have the calculation at hand, but the
probability was lower than 1:1.000.000, if I rember it correctly.
To understand the number just imagine, that the match goes
over 5 sets, the first player wins all points in the 1st, 2nd and
last set, and looses all points in the 3rd and 4th set. The 2nd
player, although playing above average still looses.

With kind regards
Marlowe

[barmar]

Could we please keep this discussion in one place? htfanclub posted the link to the thread that was started just a couple of days ago.

[ArcLight]

Sorry, I didn't realize this question was recently asked in another folder.

Thank you for the link jtfanclub