655321, on Feb 7 2008, 12:15 PM, said:
jdonn, on Feb 6 2008, 05:57 PM, said:
Cascade, on Feb 6 2008, 05:54 PM, said:
I understand.
I am trying to think how best to do a simulation here.
It depends on how often partner will push to the five-level with rubbish.
I will look at it some more later.
If partner is bidding 5
♥ on hands like the actual one, which is beyond sick, then I'm sure pass will be the winner.
Yes, it seems that you need to do 2 simulations.
First, a simulation to convince yourself that routinely bidding 5/5 is a big loser.
Second, use the results in your second simulation, such that partner's 5/5 is for real.
Ok i have done some simulations.
They have the usual problems most importantly that I used double dummy analysis etc.
1. With opener having at least five spades and exactly three hearts and approximately 12-15 hcp (less with more distribution), responder having at least five hearts and approximately 11-15 hcp (also less with more distribution), the overcaller having seven diamonds or six diamonds and a singleton and around 6-9 hcp and the advancer having four diamonds (required if you want to guarantee a ten-card fit).
I further assumed that we would double them in 5
♦ (from one hand or the other) if we did not bid 5
♥ (or higher). Initially I assumed that responder with a near minimum would never bid slam.
Under these conditions bidding 5
♥ came out at a slight loss (-0.087 IMPs per board).
On the other hand if partner always bid slam when it made then this changed dramatically to +3.27 IMPs per board.
The truth probably lies somewhere between these. Partner will sometimes be able to bid slam but will also sometimes go down when he does so.
It looks to me like a small gain for bidding 5
♥ with any minimum with three hearts.
2. Under similar conditions to the above but with responder having the exact hand given.
Now slam made nearly half of the time. Given that 5
♥ was sometimes already going down this meant that bidding slam was indicated. In the simulation there was a gain of about 2.066 IMPs per board. The variance was very large though.
If you don't believe the simulation in 1. and therefore expect partner to bid 5
♥ less often then of course slam would be even better.
---------------------------
One problem with bidding 5
♥ in the first simulation and bidding 6
♥ in the second was that you might get doubled.
When I put in a condition like you will get doubled if you are going more than 2-off then 5
♥ was still a big winner +2.752 IMPs per board but bidding 6
♥ lost most of its gains was only +0.732 IMPs per board (with an even bigger variance than before).
My conclusions from this admittedly brief and probably inadequate study is that you should probably bid 5
♥ more often than most posters think and bid 6
♥ more often than I think.