[TMorris]

bluenikki, on 2022-April-01, 10:18, said:

Under what circumstance does ace request unblock?

for me the King would ask for unblock (or count).

[smerriman]

P_Marlowe, on 2022-April-01, 00:28, said:

Hi,

if you want to set up the 4th spade, lead low otherwise high.
if you have all (most of) the combined strength, i.e. besided AK add.
values, A will be safer, you get a look, and you can continue with a
low one, you dont expect partner to get in.

With kind regards
Marlowe

PS: The given Bird / Anthias hand is a hand, that would fall, you have most
of the strength, i.e. high.

I ran a couple of sims of my own.

To keep things extremely basic, I assumed any balanced 15-17 HCP for South, and any 10-15 HCP for North, either 4333 or 4 card major; I assumed no singleton as well.

My results for the original hand over 1000 sims were:

SA: beats game 26.9%, averages 4.13 tricks
S4: beats game 15.7%, averages 3.86 tricks

Obviously all of these are a lot lower than their stats since I assume they included some 9 HCP acceptances as you would expect from a human - can rerun with more accurate acceptances if someone wants to provide a good definition - but it shows a similar advantage for SA.

If I then replace the club ace with the club 2:

SA: beats game 14.5%, averages 3.40 tricks
S4: beats game 7.3%, averages 2.86 tricks

With the latter, weaker hand:

- on 74 occasions a high spade beat game and a low spade didn't
- only on 2 occasions a low spade beat game and a high spade didn't

- on 316 occasions a high spade took more tricks
- on 22 occasions a low spade took more tricks

Again, double dummy blah blah - there were 32 cases when partner held Jxx(x), dummy held the Q, and dummy/declarer held the T, where low may cause declarer to misguess and give us an extra trick. But I'd be surprised if double dummy advantage is going to provide such a big swing on the remaining holdings - there are very few where low gains anything, and many where it costs.

[mycroft]

"In before smerriman": -) "Ace or something else?" is one of the worst double-dummy sims out there (and Bird and Anthias point this out about page 4). Because, double-dummy, if another lead is better, it usually still wins after "ace-and-switch", which they will always find.

Not saying there's anything wrong here with the analysis, necessarily. Just that that needs to be in one's mind.

Spoiler

No, I'm not upset about your - very correct - comment on the other thread. Just amused at the parallels.

[smerriman]

mycroft, on 2022-April-02, 10:57, said:

"In before smerriman": -) "Ace or something else?" is one of the worst double-dummy sims out there (and Bird and Anthias point this out about page 4). Because, double-dummy, if another lead is better, it usually still wins after "ace-and-switch", which they will always find.

You didn't get in before me though, since I beat you to it in my post in page 1

In this case I don't think there are too many parallels though - it's true that if another lead is better, double dummy will find the switch, which means you can't compare the ace to another suit lead.

But most importantly, here you're comparing it to the same suit - if another suit is better a low spade will be virtually always at least as bad (cases where you still have time to find the switch later are cancelled out by double dummy picking that up on the low lead too).

So the comparison between the two, while still having a few flaws (like the case I mentioned above, though it's too rare to affect the results on its own), is a lot more valid - and the ace at least gives a human some chance at finding the switch, compared to the zero chance of leading low.