Join Bridge Winners
All comments by Eugene Hung
1 2 3 4 ... 57 58 59 60 61 62 63 64 65 ... 69 70 71 72
You are ignoring the author of this comment. Click to temporarily show the comment.
That is the book I mentioned earlier in this thread that failed to compensate for simple constraints. From what I have seen of the book, it needed peer review from people who can properly evaluate simulations before being published. Maybe the book has started people thinking outside the box, but the authors draw too many unwarranted conclusions that are the results of flawed inputs.

For example, good players don't bid Stayman with 4333. To not account for this is obviously going to bias results towards major-suit leads, often significantly so, as Charles discovered.

Similarly, they end up recommending leading a dry ace from Axx, instead of a 6-card diamond suit. This is a result of the simulation bias of holding the lead to find the killing continuation. Using their methodology, I can get similar results on my local machine for leading Axx or Ax against 3NT, but change the ace to the king and you'll see the lead success plummet, because now you are no longer guaranteed to retain the lead to find the double-dummy continuation. Basically, their method allows “God”-like defense after trick 1, and their declarer never misguesses, so any lead that gives you the best chance to “be God” is going to look a lot better than a lead that doesn't but creates more uncertainty for declarer. Note that the misguess bias is less critical for AKx(x): when missing the AK of a suit, declarer rarely has a guess to develop tricks.

So, if looking to improve your opening leads, I would recommend NOT taking this book too seriously. Thinking is good, but playing a lot and discovering what works and doesn't work in practice is far more important.
Aug. 16, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
Although I'm a top spade advocate, there is another simulation bias in play here. I assume that you are checking to see if the contract is settable after a top spade lead, vs. after a low diamond. When you lead a top spade, you are guaranteed to hold the lead. That means you will be able to find the double-dummy killing shift 100% of the time, provided that the top spade lead itself didn't blow the contract. It usually doesn't, as supported by the sim data, as it only cost 29 times in 235 meaningful cases, around 13%.

I do think in practice you will be able to find the double-dummy killing shift a high percentage of the time, because you will see 26 cards and get a signal from partner. It's part of the reason why I think leading a top spade is such a winner. But it's not 100%.

One other tweak simmers can do is to also allow 5422 patterns with a 5-card minor (and with no worthless doubletons). I have found that opening 1NT with those patterns and good 14-bad 17 HCP works very well by avoiding nasty reverse rebids. Some people even open on 6m322 but it's not as common and harder to model with a sim.
Aug. 15, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
Sorry if I was a bit terse. But I've run a lot of simulations on my own time and I've noticed the 4333 adjustment is important in the 1NT - 3NT auction: any simulation that does not cater to that is going to give significantly biased results towards the majors. So be careful in drawing conclusions from crude simulations – apparently minor details can have large effects on outcomes. (A certain book on the market which purports to shed light on opening leads via simulations doesn't even do this easy 4333 check.)

Anyway, another good question is : how often are you likely to hold a hand with just 5 cards in the majors and the auction going 1NT - 3NT? I don't think Andrew's example situation (strong 3-card major vs. broken 5-card minor) comes up very often. You are far more likely to hold at least one 4-card major on this auction.
Aug. 15, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
The sim is flawed because responder should bid 3NT with 4333 pattern.
Aug. 15, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
The major-suit bias is not necessary. That is focusing on a minor advantage when the major advantages were covered in your article.

I made a standing even-money bet with one of my partners that high from AKx(x) is one of the best leads you can make against notrump. It's amazing how often it is the killing lead, or it helps you find the killing shift (high from KJTxx to smother dummy's singleton queen, or getting a signal from partner), or it helps partner with the defense because you place 7 HCP for him. Against fanatical AKx(x) leaders like myself, it even helps partner with a negative inference when I make a 4th-best lead and the AK of a different suit are missing – now he knows declarer has at least one of them.

So far, I've only run into one hand where I regretted leading high from AKx(x). There are many hands where leading high is break-even, but few where you wished you had led the long suit on the go. My partner has already called off the bet, because he realized he'd be going broke. If you don't believe me, try the bet yourself. Nothing focuses attention like losing (or winning) money.
Aug. 15, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
Defense is hard. You rarely get the double-dummy result from defending, especially against a low-level contract when the opponents choose not to run.

Pass, gunning for the vul opponents when you don't even know partner's second call, is a unilateral guess. It's like trying a KJ finesse in the play at trick 2 – you might be right, but you rate to do better by getting more information. You're staking the entire board on “knowing” that it's not your hand to declare. I'm not good enough to know at this point when partner has just 4 diamonds and when partner has 5-6 (so that +400 or higher is possible). Better to make a call that shows my values and lets partner know about our diamond fit, and to start hunting later when you have more information.
Aug. 15, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
The % of taking 9 tricks when partner accepts is not as important as the % of taking 9 tricks when partner rejects. Compared to just bidding 3NT, the invite only changes the contract when partner rejects. To simplify the analysis, let's say that the invitation sequence never changes the # of tricks taken. So, when you invite and partner accepts, your score is exactly the same as if you had just bid 3NT. On the other hand, when you invite and partner rejects, you win matchpoints every time partner takes 8 tricks or less, but you lose matchpoints every time partner takes 9 tricks or more. So the percentage of 9 tricks when partner rejects the invite is far more important.

If 3NT is succeeding 50% of the time overall and roughly 33% of partner's hands are accepting, with a 75% success rate, then the 67% of hands which reject are succeeding at a roughly 38% rate. So we are winning matchpoints 62% of the time and losing matchpoints 38% of the time, so inviting is the winning matchpoint strategy. If partner is accepting on even more than 33% of hands (say, 50%), then we are winning even more frequently (over 80% of the time!).

Now, in real life, inviting and accepting does not always lead to the same result as blasting game. The defense now has more information, will tend to lead more passively, and will know that partner's range is non-minimum. So in practice, I expect more invited games to go down than blasted games. However, a 12% edge is significant enough to me to overcome the greater information leakage I'm giving.
July 31, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
Apologies, I somehow thought I had the Q when constructing that “cold game”. That was a wretched 11 opposite which we wouldn't even reach game after opening, yet game had play.
July 31, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
Assuming standard opening style (11 HCP with 9 cards in two suits ok, but not 11 HCP with 8) and a 15-17 NT, I'd bid 2NT at MP and bid 3NT at IMPs. The reason I invite at MPs is because we almost certainly lack an 8-card fit. Partner doesn't have 4 spades, 6 hearts, or 5 clubs, and while 2542 is possible, it's not likely. I personally don't like rebidding 1NT with that pattern without lots of stuff in my doubletons, so even when we have an 8-card diamond fit, the hand rates to play poorly (KQ KJxxx Jxxx Kx). When our best fit is a 7-card fit, I like staying low at MP, where the plus score is king.

Why bid game at IMPs? Even though I'm not optimistic about our legitimate chances, opponents don't defend and lead perfectly, and inviting only wins significantly when partner rejects AND makes exactly 8 tricks. Even when partner is max, inviting hurts if your invite tells the opponents to lead passively (which looks like the winning strategy on this hand with no source of tricks for your side), or when it narrows partner's range to a non-minimum.

Regarding 3NT vs. 4, saying your texture (aces and spaces) may lead to 4 playing better is a misapplication of a general concept. On this specific hand, we likely have no fit. No contract rates to play well, so we might as well contract for the 9-trick game instead of the 10. Opposite the pure, prime, Ax KQJ10x Kxx xxx, you want to play in the laydown 3NT for 9 tricks, not 4 (needs a 3-3 break). 4 is better than 3NT when the trump suit gives you the tempo to set up slow tricks, but your lack of spots and fit should tell you that you don't have any slow tricks! Even KQJ KQJxx xx 10xx needs a 3-3 spade break or some squeeze to make 4, but 3NT only requires the opening leader not to lead a 5-card minor, or to have not have the A. And if partner's hearts are ratty (Jxxxx), you really don't want to be in 4.
July 31, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
When 80%+ of polled people are voting ALWAYS OPEN, it's not light, it's an opener.

Yes, opening 1 _may_ lead to a rebid problem, but it may not. Partner may raise hearts or bid spades, and when you find a major-suit fit, your hand is clearly worth an opener (6.5-7 adjusted losers). Or, the opponents may bid, and I don't think there are too many rebid problems in competitive auctions because holding both majors. If you had the J instead of the 8, almost everyone would open the bidding, rebid be damned, and yet your offensive prospects aren't much better.

If anything, I feel PASS leads to more “rebid” problems by not showing your suit and playing strength when it's convenient to do so. Heck, opposite the perfect flat 11-count (AJxx Axx Qxx xxx) game is practically cold. Good luck figuring that out after starting things off with PASS, though.
July 30, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
You even have 2 Quick Tricks. 0.5 for K, 1 for the KQ, 0.5 for the K. I've found it's okay to open hands that are less than the rule of 22, but can't think of a hand which satisfies the Rule of 22 which I would pass.
July 29, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
Fixed.
July 26, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
I had the vugraph audio open on Sunday and I didn't hear anything “astonishing” from Larry. If anything I felt Larry was being rather diplomatic while answering some of the questions that were being thrown his way.
July 23, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
That's a ridiculous analogy. Most people don't carry plastic explosives around, and if they did, it would not be for an innocent purpose.

A far better analogy is: Would you have the “guts” to report a friend whom you suspected was carrying too much toothpaste onto a plane? Toothpaste can be used to make explosive devices, but most people carry it onto a plane for innocent reasons. Similarly, cellphones can be used to cheat, but most people would like to carry them in for innocent reasons. Banning people from bringing cellphones into the playing area is like banning toothpaste from airplanes completely. There has to be a line between complete security and complete freedom.
July 23, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
Not to mention, checking in the phone costs money. Any economist will tell you, people respond to financial incentives. People also generally don't like change that restricts their freedoms, even if a restriction is justifiable. It baffles me why the ACBL not only instituted a change that did that, but also charged people for a reasonable solution. Maybe the money goes to charity, but a mandatory donation isn't a donation, it's a fee. I'd be much happier with a cellphone ban if there were someone holding phones close to the playing area without a fee (with an OPTIONAL donation). That would give off the impression “we're sorry to have to inconvenience you” which would go over much better than “these are the rules, pay up or get rid of these phones.”
July 21, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
In Standard, 2NT is definitely _not_ unusual. It shows a hand too strong to double and bid 1NT. In this specific situation, you will have a balanced 18-20ish point hand more frequently than a hand with both minors _that wants to compete_. Most often, when you have both minors with average strength and it goes 1M P P to you, you want to be passing, because the opponents likely have a fit, and possibly a game, in the other major.
July 18, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
I'm not sure how an extra 5 got into the hand, it wasn't there when I reviewed it earlier. Maybe it's a left/right (direction of your choice) -wing conspiracy!
July 17, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
At the start of the second page, I wrote “rotated to make South declarer”.
July 15, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
Mike, I wrote:

“Under this style … both partners need to understand that an overcall of a preempt can be made on a high-card light, offensive hand like this one.”

and

“This style will negatively impact your slam bidding on some hands, but guess what? Those hands don't come up nearly as often as the game hands where it's so important to find your spade fit.”

Obviously this style obviously doesn't work if your partner isn't expecting it. But when partner is, the expected value of bidding is greater than passing. This style is not guaranteed to give you positive IMPs, or 50%+ on every hand you overcall. Preempts work! But it is going to give you more good boards than bad ones as long as partner understands the range you are showing.

When you bid 3, lots of good things can happen. You find a making spade game. You find a good spade sacrifice. You find a thin spade slam. You direct a spade lead against their 3NT or 4 contract. Your spade bid helps your partner on defense. The opponents misjudge due to the competition. You end up in a bad contract but make on bad defense.
(Most people declare better than they defend.)

Bad things can happen too: you can go for a number, your side might bid too high, you might double them in a making contract. The bad things happen less often if your partner knows this hand type is possible.

When you pass, you are guessing it's right to pass. When you bid, you are guessing it's right to bid. But when you bid, you more frequently have ways to backtrack and say “no, it's right to pass”, later in the auction. When you pass, it's much more committal and partner is going to have more difficulty figuring out what to do, because pass is so vague. Show the general shape and offensive nature of your hand (spades), not points, first, and your partnership will make much better decisions about strain.
July 15, 2012
You are ignoring the author of this comment. Click to temporarily show the comment.
The idea “No preempts over preempts” applies to jump overcalls (such as a bid of 4), not overcalls. Passing is the 1950s textbook call, but watch great players on Vugraph. You will see them overcalling on hands like these, not passing. Finding your fit is critical in competitive auctions. 3 is not risk-free, but passing faces far more risks, as it is so often right for your side to compete in spades.

Under this style, which is what I believe to be the expert modern style, both partners need to understand that an overcall of a preempt can be made on a high-card light, offensive hand like this one. Partner should not drive to slam with a 13-point hand. This style will negatively impact your slam bidding on some hands, but guess what? Those hands don't come up nearly as often as the game hands where it's so important to find your spade fit. The key factor for making a high-offense overcall is shortness in the enemy suit. With length, you don't need to stretch because then you can pass and rely on partner to balance with the points and shortness needed if you have a game. With the shortness, you need to act or partner may pass out 3 with some mediocre hand that makes game cold, like Kxx xxxx KJx Axx. (And partner could be even stronger and pass it out.)
July 14, 2012
1 2 3 4 ... 57 58 59 60 61 62 63 64 65 ... 69 70 71 72
.

Bottom Home Top