jvage, on 2019-December-13, 12:10, said:

I guess you meant the opposite, Paul. With your strategy I will always make (while I would sometimes misguess if you had chosen the opposite strategy)

I think that might just go to show that this defensive strategy would rarely be found in practice. Espescially since I didn't actually tell you how I would play (against most opponents I do however think I would guess the diamondsplit better than just the original odds). My point was just that it's close, possibly closer than the 3% difference that you mention, which is already relatively close.

Why would I not lie to you about my strategy? Yes, I agree that it is close. However, if you "know" that East has the ace of diamonds, it is not so close. What NS submitted to the referee was the following:

Without the hesitation, the normal line is to play a diamond to the nine. This succeeds whenever the ten of diamonds is onside and the diamonds are 4-2 or 3-3, Given that they break this well about 85% of the time, this line is close to 42.5%. After the hesitation we “know” that East “must” have the ace of diamonds, or he would have nothing to think about. And giving suit preference is not a “bridge reason” for the BIT.

The new possibilities are therefore East having Ax, ATx, Axx, Axxx, ATxx, and ATxxx. In the last two he would have nothing to think about as he would never play the ace in those situations, and he would probably have led a diamond from ATxxx in any case. However, we shall just consider that he has the ace and no more than four diamonds (as it is not relevant what we play if he has five). The relative odds:

Ax 6.4% (note that AT doubleton has been eliminated from the enquiries)

ATx 7.2%

Axx 10.8%

Axxx 6.4%

ATxx 9.6%

We have eliminated some holdings, as being impossible, and we “know” the ace of diamonds is with East. Of course the above has to be scaled so that it adds up to 100%

That gives

Ax 15.8%

ATx 17.8%

Axx 26.7%

Axxx 15.8%

ATxx 23.8%

So, at the point where the critical decision has to be made, playing the jack works in 44.5% of the cases (we are never picking Ax) and playing the nine works in only 41.6%. It would be a clear error to play the nine, and two very strong players whom I asked played the jack when given the problem (with the hesitation).

**If one considers, as I do, that East is unlikely to hesitate with ATxx (as he would never think of playing the ace with that), then I think finessing the nine drops to 17.8% - a ridiculous line.** I do not pretend for one moment that North did any of the maths at the table, but she did originally intend to play the nine, thought for about ten seconds when East broke tempo, and then judged (correctly) to play the jack. She called the TD and the BIT was agreed immediately. I submit that this ruling is truly awful and not in the ballpark and 100% of 3NT= is correct, which is what we are seeking.

In this example, the play by East of hesitating without the ace should have attracted a procedural penalty. Instead, East benefited by deflecting North from the winning line. I agree that he “was not aware”, but that does not matter. He “could have been aware”, especially looking at Txxx in diamonds!

Now I agree that one might be able to do a bit better by combining chances if the jack wins, as one does not have to play for 3-3 diamonds. However if the nine forces the ace, you are very likely to make. If you play the jack and it holds, then you are probably going to play for diamonds three-three rather than playing for two heart tricks. All this is irrelevant of course; an innocent person drew a false inference from an opponent's infraction, so the TD should adjust.

I prefer to give the lawmakers credit for stating things for a reason - barmar