BBO Discussion Forums: Unlucky expert - BBO Discussion Forums

So I had a bad session yesterday at Biarrtiz open pairs, facing a lot of decisions that I mostly made wrong.

This is one example:



The lead was ♠5 (4th best, low promises a honnor), and East produces ♠J

Spades probably are 5-1, since leading from AQ95 is a bad idea, specially when she likelly has a side 4 card suit.

Winning ♠K relies on ♦KQ being onside (25%), while ducking relies on ♣Q being onside (50%), so if spades are really a big favourite to be 5-1, then ducking is correct.

Right?


Well I missed an important factor, if spades are 5-1, the finesses are not 50-50, with the vacant spaces 8-12 they are 40-60, so the numbers are actually like this:

♦KQ on your right: 36%
♣Q on your left: 40%

Now risking a 4-2 spade break doesn't look so promising uh?

LHO had ♠AQ952 ♥xx ♦xxx ♣9xx

If you duck, you're gonna look very silly if RHO has Jx and you end up losing 4 spades and a diamond.

Especially when you could have won by taking and relying on a diamond honor onside, which is what the field will be doing anyway.

This isn't a hand to get clever. It's just a "play along" 3NT and you should do just that. At least I think so

I'm not sure I follow the intended analysis.

Ducking will be ok if spades are 5-1 and either diamond is onside and the ♣Q as well. By the vacant space argument both ♦ are wrong .4x.4=.16, so the double diamond finesse is 84%, and the club finesse 40%, for a net of 32%. That compares to 75% x 50% =37% when there is no vacant space consideration, so argues slightly for winning the diamond. Is that the idea?

Winning the opening spade lead here works whenever both diamonds are onside, 35% plus or minus depending on what vacant space assumption you make, so already it's nearly equal with even the optimistic 37% for ducking. Add even a 5% chance of spades 4-2 with diamond honors split and Bob's your uncle.

I like the argument that spades are 5-1 but surely it has to be mitigated by the huge dealing odds in the opposite direction. If LHO leads at random from equally long longest suits, the spade suit would be 70% 37% to be 4 cards. I can see halving that based on judgment about leading proclivities, but cutting it to zero? Ducking the first trick is taking an extreme position against even a small chance of ♠4-2.

(Edit - on reconsideration the actual spot led increases the chance of 5-1 spades.)

This post has been edited by ceeb: 2014-July-07, 13:13

Maybe my attitude towards this hand is part of the explanation for why I generally do better at imps than at mps, but I just don't get the notion that it is a bad idea to lead away from AQ95 in this auction.

That 9 is a huge card. AQ75 is a bad holding from which to lead, because you need a lot of luck to be able to run 3 or 4 tricks, but that 9 can really grow up opposite lots of layouts, including this one.

Meanwhile, after 1N 3N, we are leading a 3 or 4 card minor? From say Hxxx? And that is safer? Or do we lead from xxx or xx, in the latter case maybe picking off partner's holding?

Passive leads can be winning leads at mps, but it is far too easy to over-rate that notion.

This is important here because, against me, it would be foolish to infer that spades are 'probably' 5-1. They might well be, but I surely wouldn't recommend the cute duck here.

I think it absolutely clear to win, cross in hearts, diamond to the J, and later cross back to dummy, assuming we aren't yet down, and lead the diamond 10. I truly don't think it to be close at all.

I found this hand very interesting, mostly because my thought process was similar to Fluffy's (I ignored empty space implications on the finesses). Yes, it does not make a huge difference, but it makes a reasonable difference and in a close decision that can be a factor.

This is the type of hand where other factors should sway your decision. For instance, a quick spade lead makes a 5 card spade suit more likely against almost everyone (I don't think anyone even if they would lead from AQ95 would be thrilled about it, it is usually that they have no other good lead, like Jx of hearts and some minors). Conversely, a slow/tortured spade lead is probably a 4 card suit. Even a normal tempo spade lead is very likely to be 5 I think.

Also, whether this is a qualifying game or a finals would influence my decision heavily, especially in an event with no carryover and not a massive cut, I think not many people are going to duck the lead here so it's a big play and if my goal is just to qualify and I think I'm likely to do that, I don't think there is enough of an edge in it to risk the variance. I always post here that people are WAY too into what is the field doing and blah blah but this is a spot where it would impact me since I think it is a normal auction (unless it's a heavy weak NT field), and if they are leading from a 5 card suit it's very normal, and I think I can accurately predict that few declarers will duck (if it's not a normal lead then they have a 4 card suit, which is a good time to win it!), and I think it is already a pretty close decision either way.

I really think it is possible to get a lot from the tempo of the lead vs most people, that would probably sway me the most. But I'm sure in a qualifying game I would also always win the K unless I was a bad pair. The point about empty spaces, and maybe judging based on your opps age/nationality/skill level how often they would lead from AQ95 is also important, obviously against a pair who just knows 4th from your longest and strongest a 4 card suit becomes much more likely again.

I don't think you made a horrible play or anything Fluffy, and I do think the point about empty spaces is easy to overlook and might (or might not) change your decision.

@Mikeh, I think even at imps AQ9x is not great, it at best sets up 1 trick and if your partner has two tricks a non spade lead might beat them anyways. Obviously that is an overly simplistic way to look at it though. On the other hand, if your partner has 1 trick and you went from 4 spade tricks to 3 that's a disaster. But of course this is a matchpoint event and I think AQ9x is really really unappealing there.

PhantomSac, on 2014-July-07, 18:40, said:

My experience is obviously different from yours. Especially at imps. Which is why the game is so much fun, but opening leads are, I think, one of the stronger areas of my game, such as it is (or was).

mikeh, on 2014-July-07, 20:24, said:

Yeah sorry I did not mean to come across like I know what the right lead is, I think basically no one knows much, and I certainly don't I was just telling you my reasoning for in general not liking that lead at imps, I think it is an interesting problem in that case. I think many world class players would lead a spade at imps and many would not, partially evidenced by this thread heh.

I don't even know if it is an experience based thing for me, I have never been a spade leader with that type of holding but it is more because of what I have thought not what I have done. My dad and some other partners I have had like Hamman have definitely led spades at imps in those situations, sometimes it works and sometimes it doesn't heh.

I do feel like at MP AQ9x would be a bad lead though, feel pretty strongly about it at that form of the game (though ofc I could be wrong. Just felt that your comment about imps was an interesting point to this hand worth commenting on.

edit: after re-reading your post I think you were saying even at MP you might lead a spade but you feel stronger about it at imps? In that case, I disagree with you at MP and at imps I am different from you but am not sure that I disagree; I have no idea/strong opinion heh. And FWIW I feel like I am pretty weak at opening leads in general, so take that for whatever it's worth!

This post has been edited by PhantomSac: 2014-July-07, 22:05

Fluffy, on 2014-July-05, 03:49, said:

Let's say you duck the spade jack. I only see 8 tricks. 3 hearts, 3 diamonds and 2 clubs.

jogs, on 2014-July-07, 22:03, said:

You are hoping to score a club finesse if you duck. If you win you are hoping spades are 4-2 or the KQ of diamonds are on. I think fluffy's point is that the club hook is more likely to be off and the KQ of diamonds are more likely to be on if LHO has 5 spades and RHO has 1 and that changes the odds. Ofc if you think spades have a reasonable chance of being 4-2 winning is clearly right... That is a different story (and Mikeh's views are telling that even if you are a non spade leader at MP with AQ9x, some very good players are, so winning might just be clearly right regardless of the empty space thing.)

Fluffy, on 2014-July-05, 03:49, said:

I don't agree with your calculations. The location of the ♦K and ♦Q are not independent. One can't just multiply .6*.6 to determine its probability. It is from a hypergeometric distribution.

The denominator is 20C2. 20*19 / 2*1 = 190. 190 ways. ♦KQ both being with East is 12*11/2*1 (=66). That works out to 34.7%

The ♣Q on the left still fails when LHO has both the ♦K and ♦Q. ♦KQ both with LHO is 8*7/2*1 (=28). Therefore it is 40% of 190-28 (=64.8).

Winning the first trick is a small favorite when spades are 5-1.

Your chances of making 3NT legitimately are not very good, so at MP's there is a good case for winning the spade and hooking the diamond simply hoping to stay even with the field.

I strive to be a proper Bayesian, but you can't realistically assign spades as being AQ95 opposite J2 as having zero probability.

Rethinking all calculations.

♦KQ both with East is 12*11/2*1 (=66).
♦KQ split between East and West is 12*8 (=96).
♦KQ both with West is 8*7/2*1 (=28)
66 + 96 + 28 = 190 ways

Winning the 1st trick works whenever ♦KQ both with East.
That's 66 out of 190 ways.

Now using vacant spaces for ♣Q.

♦KQ both with East is 12*11/2*1 (=66).
8 spaces in West and 10 spaces remaining in East.
66 * 8/18 = 11 * 8/3 = 88/3

♦KQ split between East and West is 12*8 (=96).
7 spaces remaining in West and 11 spaces remaining in East.
96 * 7/18 = 16 * 7/3 = 112/3

Multiply by 3 to rid ourselves of fractions.
190 ways becomes 570 ways.
88 + 112 = 200

There are 200 out of 570 ways for ducking to succeed.

66 * 3 = 198
There are 198 out of 570 ways for winning the 1st trick to succeed.

1st conclusion: given that spades are 5-1.
Ducking a razor thin favorite (200vs198) over winning the 1st trick.

Final conclusion: which includes chances of spades being 4-2.
It must be correct to win the 1st trick.
---------------
Of course none of us can make these precise calculations at the table.

Flat 11 opposite 14. Will a part of the field be willing to play this board at 1NT? We better maximize our chances of making this hand.

If I duck the spade, can't i make it very often getting 3 diamonds and 3 hearts and two clubs and one spade?

SUppose I duck the spade, win the return of a heart and play a low diamond to the 9. If lho wins this and returns a spade I have 9, so he returns a heart, so I win in dummy and run the diamond T, and then take three diamonds pitching a spade from dummy. Now i can cash the heart and play AKJ of clubs, now i win with the club finesse right or Qx offside. Right?

phil_20686, on 2014-July-20, 14:22, said:

It depends what you mean by "very often". Against best defence, that requires:
- Spades 5-1
- At least one diamond onside
- Either ♣Q onside or LHO's being 5xx5.

gnasher, on 2014-July-20, 14:57, said:

+ defensive error.

Its not guaranteed that west will make the right discards on the 7 red suit tricks. Especially if he is 5422 with Tx or 9x of clubs. In practice you might make on this kind of layout.

Or if lho has T9 of clubs you should always make I think. You cash the four diamonds, and the three hearts and he has, e.g. AQ9xx xx Kxx T9x he can afford one spade, but then has to pitch a club, and now a top club and running the J will make whenever lho has the club Q and whenever he has T9x, as either he is down to Qx or you are pinning the T9.

So it seems like there are lots of ancillary chances if you duck. If lho heroically forsees this and switches to a club, well you just put up the J and lho will still get squeezed/endplayed if he has the T9 of clubs.

So ducking wins: whenever the club Q is onside, or whenever the clubs are 5-2 or whenever lho holds the T9 of clubs.