[masonbarge]

(This is an actual hand from a Bot game.)

Matchpoints.

---

W leads ♥A followed by ♥Q, ruffed in hand. Assume the Bot would bid 4♥ with a void or 2 singletons.

[nate_m]

masonbarge, on 2015-February-12, 16:12, said:

(This is an actual hand from a Bot game.)

Matchpoints.

---

W leads ♥A followed by ♥Q, ruffed in hand. Assume the Bot would bid 4♥ with a void or 2 singletons.

Cross in diamonds, run the 9 of spades is my instinct.

Let's see how it plays out...

I think that West has 7 hearts and no stiff minor, else they play it at trick 2. I ignore the HCP requirements for preempts because in my experience they aren't that accurate. Therefore, the probable shapes and the number of combinations associated with them as I see it are:

Shape
1732: 1050
1723: 1750
2722: 2100

Running the 9 of spades wins in 3580 cases (picks up the 4-1s that aren't stiff J and 3-2s with J onside) and playing low to the KQ (or low to the K then finesse 10 if it loses, as I don't recall robos ducking from Ax, so we'll assume that works) wins in 2370 cases (stiff J offside plus all 3-2s but AJ doubleton off). Under these highly restrictive assumptions running the 9 is percentage.

Of course, my assumptions are far from 100% but that would be my line.

[masonbarge]

nate_m, on 2015-February-12, 18:13, said:

Cross in diamonds, run the 9 of spades.

I believe that is not optimal.

[wanoff]

♠Q and if West follows, cross in diamonds at first opportunity and run ♠9.
Only a problem now if trumps 4-1, so shorten trumps if necessary to match Easts length, cash clubs and go for the trump coup.

[masonbarge]

wanoff, on 2015-February-13, 04:32, said:

♠Q and if West follows, cross in diamonds at first opportunity and run ♠9.
Only a problem now if trumps 4-1, so shorten trumps if necessary to match Easts length, cash clubs and go for the trump coup.

Almost there for what I see as optimal play.

East goes up with ♠A and returns a low club.

[nate_m]

masonbarge, on 2015-February-12, 18:44, said:

I believe that is not optimal.

It's far from obvious to me that playing a top spade first is better.

If trumps are 1-4, hooking on the first round picks up any stiff but the jack.

If trumps are 2-3, playing a top spade first doesn't help, as when it loses you intend to finesse E for the jack of spades anyway, so you still lose to any doubleton with the jack.

If trumps are 1-4 and you bang a top spade first, you pick up stiff jack offside (360 layouts). However, after the top spade loses and you win the return and run the 9 getting the bad news, you need 3 rounds of diamonds to stand up for the trump coup to operate. On 840 layouts (West 1732 without stiff jack) you will now fail, whereas running the 9 would have succeeded.

The preceding analysis assumes West would have shifted to a minor suit singleton if they had one, which is not IMO a totally crazy assumption to make.

I appreciate hearing that you do not believe my line to be optimal. What line do you prefer, and why do you prefer it?

[PhilKing]

I would just cross to a diamond and run the ♠T. This just pays off to West being 2722 with ♠Jx, which I don't think is that likely.

[suokko]

We have only two top losers. So we can just play trump from top from our hand and avoid LHO getting in with ♠jack to force us or any unlikely ruff for LHO.

EDIT: IMP mode left on from yesterday team match.

So yes. MP finessing spade jack is something that I would have to do. 4-2-?-? for RHO means trump coup works when RHO has 4-2-3-4 or 4-2-4-3 (LHO 1-7-2-3 or 1-7-1-4) but fails with RHO 4-2-(25) or 4-2-1-6 (LHO 1-7-3-2 or 1-7-0-6 or 1-7-4-1). Which leaves us losing option if 1-7-3-2 opener doesn't have trump jack. That makes more combinations won if crossing to dummy with diamond and finessing first trick. Of course that could go down too if LHO has Jxx AQJxxxx xx x and managed to pull a diamond ruff after our play. But the shifting to singleton club would have worked for defense too.

But there is alternative to play spade ace for RHO and play twice from dummy to honors. That wins any 3-2 trumps with marked spade ace on side (opener 2-7-2-2, 2-7-3-1, 3-7-2-1 making that more likely than 1-7-(32) and 1-7-(41) shapes for 4-1 trump break. That line even picks the singleton jack if trumps are 4-1.

[masonbarge]

suokko, on 2015-February-14, 06:08, said:

We have only two top losers. So we can just play trump from top from our hand and avoid LHO getting in with ♠jack to force us or any unlikely ruff for LHO.

EDIT: IMP mode left on from yesterday team match.

So yes. MP finessing spade jack is something that I would have to do. 4-2-?-? for RHO means trump coup works when RHO has 4-2-3-4 or 4-2-4-3 (LHO 1-7-2-3 or 1-7-1-4) but fails with RHO 4-2-(25) or 4-2-1-6 (LHO 1-7-3-2 or 1-7-0-6 or 1-7-4-1). Which leaves us losing option if 1-7-3-2 opener doesn't have trump jack. That makes more combinations won if crossing to dummy with diamond and finessing first trick. Of course that could go down too if LHO has Jxx AQJxxxx xx x and managed to pull a diamond ruff after our play. But the shifting to singleton club would have worked for defense too.

But there is alternative to play spade ace for RHO and play twice from dummy to honors. That wins any 3-2 trumps with marked spade ace on side (opener 2-7-2-2, 2-7-3-1, 3-7-2-1 making that more likely than 1-7-(32) and 1-7-(41) shapes for 4-1 trump break. That line even picks the singleton jack if trumps are 4-1.

I think that's exactly right.

RHO might also play the ♠A on the second spade from dummy from AJxx, which is what actually happened. It's easy to sit and ponder the hand and say "of course he should duck", but it's a different matter to do it at the table.

[nate_m]

masonbarge, on 2015-February-14, 15:12, said:

I think that's exactly right.

RHO might also play the ♠A on the second spade from dummy from AJxx, which is what actually happened. It's easy to sit and ponder the hand and say "of course he should duck", but it's a different matter to do it at the table.

I mean, with the example hands suggested in the post above, it's hugely likely LHO shifts to their stiff minor, so they can basically be discounted.

Flying the ace from AJxx on the second spade from hand is a novice play, as it is successful on 0 layouts, blows a trick in practice on almost every single possible layout, and as a bonus is consistent with general principles (second hand low). Playing low is something I would expect any non-beginner to find, so I would estimate your odds of picking up AJxx on an egregious misdefense and a trump coup as near zero vs. human opponents who don't have issues following suit. You are playing against robots, so you might see the A from AJxx a fair bit of the time.

Fun fact: If the robots play the A from AJxx 100% of the time, you break even on your line. If the robots EVER duck AJxx, my line is better, if we assume opponents shift to a stiff minor.

[nige1]

masonbarge, on 2015-February-12, 16:12, said:

(This is an actual hand from a Bot game.)
Matchpoints.
W leads ♥A followed by ♥Q, ruffed in hand. Assume the Bot would bid 4♥ with a void or 2 singletons.

---

nate_m, on 2015-February-12, 18:13, said:

Cross in diamonds, run the 9 of spades is my instinct.
Let's see how it plays out...
I think that West has 7 hearts and no stiff minor, else they play it at trick 2. I ignore the HCP requirements for preempts because in my experience they aren't that accurate. Therefore, the probable shapes and the number of combinations associated with them as I see it are: Shape
1732: 1050
1723: 1750
2722: 2100
Running the 9 of spades wins in 3580 cases (picks up the 4-1s that aren't stiff J and 3-2s with J onside) and playing low to the KQ (or low to the K then finesse 10 if it loses, as I don't recall robos ducking from Ax, so we'll assume that works) wins in 2370 cases (stiff J offside plus all 3-2s but AJ doubleton off). Under these highly restrictive assumptions running the 9 is percentage.
Of course, my assumptions are far from 100% but that would be my line.

I like nate_m's line, especially given his assumptions. Suppose you relax those assumptions. Now the contract is sometimes defeated, when LHO is short in ♦. For example
♠ x x x ♥ A Q J x x x x ♦- ♣ x x x
♠ x x ♥ A Q J x x x x ♦- ♣ x x x x
♠ J x x ♥ A Q J x x x x ♦ x ♣ x x
♠ J x ♥ A Q J x x x x ♦ x ♣ x x x
For example, LHO wins ♠J, RHO ruffs a ♥ with ♠A, and LHO ruffs a ♦.

In some such cases, It's safer to cross to dummy, by overtaking ♣Q. If nate_m's assumption fails and LHO is short in ♦s, then declarer might recover from his profligacy in ♣s, by a trump squeeze in the minors. In the diagram position, declarer cashes his penultimate trump, discarding ♦6 from dummy. Now, when he cashes ♦AK, he will know how to continue

---