[nullve]

nullve, on 2017-December-25, 10:20, said:

The Woodson 1N opening (which shows a balanced hand with either 10-12 or 16-18 hcp) is not great when Responder has something like 8-11 bal. without a 5-card major. But suppose that with this hand type Responder, instead of always bidding (just in case Opener has 16-18, but at the risk of having to play 2N on a combined 18-23 count), tries to guess whether partner has 10-12 or 16-18 based on RHO's reaction to the 1N + alert. Would that be a form of coffeehousing, or just on a par with playing the jack from KJTx in hand opposite A98x, waiting for LHO's reaction before deciding which way to finesse?

My very small sims have suggested that if the bidding goes

P*-(P)-1N**-(P),

* 0-10 if BAL
** BAL with either 8-10 or 14-16 hcp

then Opener will nearly always have 14-16 hcp even though points could easily be divided rather evenly (say, 10-8-10-12) around the table.

So, assuming Responder can now more or less pretend partner has opened a 14-16 NT (well, using the normal NT structure might not be a good idea), maybe this Woodson-like 8-10 or 14-16 NT in 3rd seat is "playable" in the more normal sense where Overcaller's body language is ignored?

[awm]

This sim will depend a lot on which assumptions you make. In particular:

1. When 1st seat has a very weak hand the odds of partner having 14-16 must be very high, but more interesting is the case where 1st seat has 9-10. Now the odds will be different.

2. Do opponents play X for penalty and if so what is the minimum strength? If opponents routinely pass all balanced hands in 4th here the odds that 4th chair has cards (and 3rd chair is weak) will be much better than if opponents double for penalty aggressively and yet passed here.

3. Also problematic is the case where 4th hand overcalls. 1st chair will often have a competitive hand opposite 14-16 but want to stay out opposite 8-10?

[Vampyr]

You can take inferences from opponents’ behaviour at your own risk of course, as long as you do not play the card for a two-way finesse and leave it on the table for some time waiting to see who reacts. This is cheating, whether you are an English international or not.

I have nothing against the 2-way NT you are considering, but I do not recommend opening a potential 10-12 in third seat, at least at teams.

[nullve]

awm, on 2020-June-15, 05:48, said:

1. When 1st seat has a very weak hand the odds of partner having 14-16 must be very high, but more interesting is the case where 1st seat has 9-10. Now the odds will be different.

Result of sim generously assuming North and East would have opened the bidding with 11+ hcp, that only classical balanced shapes ((4333),(4432) and (5332)) are in South's 8-10 or 14-16 1N and that West would not have passed with 13+ hcp:

Frequency :
    8	    4185
    9	   13609
   10	   27066
   11	       0
   12	       0
   13	       0
   14	   75282
   15	   74007
   16	   68130
Generated 10000000 hands
Produced 262279 hands
Initial random seed 1592742339
Time needed   19.122 sec

using the following Dealer script:

hcp(north)<11 and hcp(east)<11 and ((hcp(south)>7 and hcp(south)<11) or (hcp(south)>13 and hcp(south)<17)) and shape(south, any 4333 + any 4432 + any 5332) and hcp(west)<13
action frequency (hcp(south),8,16)

So the 14-16 NT is only about five times as likely as the 8-10 NT, which means that

nullve, on 2020-June-15, 04:36, said:

Opener will nearly always have 14-16 hcp

is already a huge exaggeration. Sorry about that.

And, yes, the odds get considerably worse when North has 9 or 10 hcp, 9-10 hcp being the problem range*.

Frequency :
    8	    4332
    9	   13516
   10	   22776
   11	       0
   12	       0
   13	       0
   14	   35804
   15	   31025
   16	   25255
Generated 10000000 hands
Produced 132708 hands
Initial random seed 1592742180
Time needed   19.080 sec

using

hcp(north)>8 and hcp(north)<11 and hcp(east)<11 and ((hcp(south)>7 and hcp(south)<11) or (hcp(south)>13 and hcp(south)<17)) and shape(south, any 4333 + any 4432 + any 5332) and hcp(west)<13
action frequency (hcp(south),8,16)

.
Now the 14-16 NT is only about twice as likely.

* Someone used to a "pass or bash" style opposite a 14-16 NT would probably happily pass most balanced 9 counts, though.