HighLow21

Good response. Thank you for this. (Yes, South declares.) I'm going to assume for the rest of this discussion that suit-by-suit probabilities are independent of each other (they are not, but it's not far off and it makes the math actually doable). Your line is better than what I tried at the table. I completely spaced out on the possibilities in spades, because I was so upset about the bidding. The primary line I was going for was 5 tricks in hearts plus 3 tricks in diamonds (6-1 or 5-2 split with one or both honors in the short hand, OR 4-3 split with both honors in the short hand). This is about 26.8% for the diamonds, I believe. Your spades option seems like a little over 50/50 to me; much better. So we are going for 7 tricks from the majors. By my count, being able to play ♥ for 5 tricks requires W to have Kxx♥ (17.8% of the time) or Kx♥ (16.1% of the time). This is 33.9% I think. Using the spade play you described, without a squeeze or error by the defense the slam is thus, I believe, half of this, or 17.0%. Add to this the possibility of dropping K♠ (either Kx♠ with East or KJ♠ with West -- by itself, 8.1% I think) adds another 2.7%. We're at 19.7% now. Let's assume normal probabilities on the club lead for simplicity (though a decent defender is more likely to pick something else with Txx or Jxx, if there was any other choice, but there's a fair chance there wasn't a good choice, especially holding Kxx(x)♥). Txx or Jxx with West originally is 17.8%, so by my math this adds [(1 - 19.7%) x (17.8%) x (33.9%)] = 4.8%. We are now at 24.5%. If W has the K♥ to 4 or more, the only way to get 5 tricks is via a squeeze--and there seem to be so many available. Which squeeze is best? And does that get us all the way up to 30%? It should be noted that at the table I estimated about 15% for the slam, but that's because I was starting with the diamond play. It's still a bad slam, but not as terrible as I thought. One consideration on squeeze play here: after the A♦ is unblocked and a heart finesse is taken, dummy will quickly become short of entries. Any clue how to figure out which squeeze play is best in light of this? And feel free to correct me on any of my math here.

Not if the 10♠ is single. I'd play for that.

The double was fine. The fact that they can make 4NT and not 4♠ is a fluke in my view. But given that it was matchpoints, it stood little to gain and a fair amount to lose. (Edited because original comment incorrectly assumed partner was on lead.)

Right, we can only talk about "good" relative to the opposition -- not to the general bridge population as a whole. Assuming a 5 for standard deviation, the typical variation over 100 boards will be 5 x sqrt(100) = 5 x 10 = 50. So +/-50 IMPs. A 0.6 average would imply 60 IMPs, which is not "statistically significant" but you typically need a ratio of around 2 or better for that --> your ratio is 60/50 = 1.2. This value is called the t-statistic, and the corresponding p-value is around 11%. (To take out the math and jargon here: if you WERE, in fact, equal, then after 100 boards, you would only be up +0.6 IMPs about 11% of the time. Similarly, you would only be down -0.6 IMPs 11% of the time.) So as a statistician, I would say that 100 boards is not enough to tell for reasonably certain. The magic "p-value" number for statisticians is typically around 5%. But it's not far off; at 500 hands, it would start to be pretty certain. (At 500 boards, the standard deviation of the accumulated IMPs total would be 5 x sqrt(500) = 112 IMPs, but you're up by +0.6 x 500 = 300 total; this is would happen by random chance well less than 1% of the time.) As a bridge player, I would say that +0.6 means that you are very likely the better pair, though I'm not conclusively convinced yet. :-) One thing I will tell you: I track many of my results in a spreadsheet and calculate the running IMPs total on the previous 200 hands at all times. (Yes, I am a math nerd.) It is very rare that, at any given time, the previous 200 hands have an average IMP score that is as much as 0.6 above (or below) my long term IMPs average. Hope this helps.

It all depends on what partner means for his double, so if you aren't sure, I would recommend interpreting as Lightner, requesting you lead your longest suit (or really, in this case, lead a minor). If it turns out a club is right, then those are the breaks. Partner's double means one of three things, only two of which are reasonable: 1. They aren't making 7♥ and he knows it. If declarer knows what he is doing, this is unlikely. 2. He's afraid if you make the normal lead of a spade they will MAKE the contract, so lead something else. 3. He's upset at how his 3♠ rebid turned out and is doubling out of anger or is gambling. If it's #1, it doesn't matter terribly much what you lead. You're going plus 95% of the time no matter what you lead. If it's #2, then it does matter what you lead and the diamond is probably the best lead under most circumstances. (1) Partner may be ruffing. (2) Partner may have the ace. Both are possible with a club lead, but (1) is more likely with a diamond lead because you have 5 of them. If it's #3, you should leave the table and not play with this partner again. In theory if it is #2 and partner is right (about a spade lead giving them the contract), you only have to guess right about 10%-20% of the time for his double to show a profit in the long run. Still, I'm inclined to distrust this partner's judgment for bidding 3♠.

Assuming that the swings are random and independently distributed, then you can mutliply the st deviation of a single board by the square root of the number of boards played to get your error measurement. I stated in a previous reply that the standard deviation of IMP swings is, in my experience, about 5. 5 times the square root of 8 yields about 14. In other words, the standard deviation of the score for the 8-board match would be about +/-14, which means, roughly speaking, about 2/3 of the time it'd be smaller than that and about 1/3 of the time, larger (in other words, -15 or lower, or +15 or higher). In 32 boards, the same math applies: 5 times the square root of 32 yields about 28. So about 2/3 of the time, the winner will net 28 IMPs or fewer, while 1/3 of the time, the winner will net 29 IMPs or more. And actually, my personal, if overly precise, standard deviation of IMPs estimate is closer to 5.3. Using this instead of 5 above, I would get standard errors of 15 and 30 rather than 14 and 28. In this case, intuition and math lead to very similar answers. :-)

Pass and expect them to make 10-11 tricks most of the time.

Here's a good article -- it claims that roughly 80% are left alone, 14% will benefit, and 6% will face a tough choice. The article claims about 3% of US residents, total, will be made worse off. I'm not saying it's necessarily right, but it's a good data point in favor of ACA. http://www.newyorker.com/online/blogs/newsdesk/2013/10/obamacares-three-per-cent.html?mobify=0

It does not properly account for them yet; as more names get added to the database, its ability to account for the skill level of the opponents will continue to improve. Hence all my caveats about its use. But all that said -- it beats the pants off of not knowing anything, and it also beats the pants off of the "self-rating" that people give. But to say that "it rewards bunny bashing," and nothing else, is demonstrably false.

In that case, 0+. If you are equally matched then a score over zero is good and under zero, not good. The standard deviation of IMPs is, in my experience, around 5. This means you should get 10+ IMP swings (in either direction) around 5% of the time and 5-10 IMP swings around a quarter of the time.

It's been facepalmingly bad.

Here's an excellent article referring to the sabotage from the right from the very start. The GOP has intentionally been doing everything to make the government-centric plan fail, because, well... they run on a campaign of "government is bad." (Except defense spending and border protection of course.) Again -- ACA would be in much better shape of the right had cooperated, rather than sabotaged. In how many situations do half of an entity's employees have to spend so much time and energy doing the work for everybody, while having to obviate obstacles put up by the other half, who get to keep their jobs? http://dish.andrewsullivan.com/2013/11/04/how-hated-of-obamacare-has-hurt-it/?share=email&nb=1

I'm not blaming Fox News for everything -- they are just trying to make money, and they can't make money without a gullible audience. They are one of the main culprits, though, of the right-wing misinformation machine is very large, very scattered, and very lucrative. See Rush, Erickson... hell even the NY Post and WSJ nowadays. There are plenty of organizations making money by giving extremely biased perspectives on everything political -- some on the left, but an absolute ton on the right. And by no means am I happy about how Obamacare has been handled thus far. I expected better on the website, I expected better on the overall management of the process, and I expected more favorable (or at least more balanced) press coverage. The administration has made numerous flaws in this rollout, not least of which on the PR side, where it effectively lied about being able to keep the insurance you have if you like it. (It can be argued that Obama, when he said this repeatedly, was referring only to insurance that would REMAIN LEGAL after the ACA took effect -- but when you have to retroactively caveat bold statements like that, you've failed the PR quality test in my view.) The problem here is that, it seems, all we are getting in terms of serious discussion of ACA comes in two flavors: 1. Criticism of the rollout/website/certain features of the ACA plan. 2. Anecdotal stories about individuals who are made worse off by the new system. On (1.) -- there will be problems with the rollout, as ACA represents a massive change in a massive industry. There should have been fewer problems or less severe ones, but they ALL (or at least *mostly*) will be fixed. (It should be noted that if the GOP spent half as much time trying to constructively improve ACA as bashing it or trying to repeal it, for over 3 years now, it would be a better law. Thankfully, the Democrats are at least interested in governing rather than only obstructing and squabbling.) On (2.) -- anecdotal evidence is interesting as it gives color to a situation, adds depth, names and faces, details. However, it's anecdotal. When a new system rolls out there will ALWAYS be winners and losers in that system. Anecdotes are a poor substitute for statistics. What I'd like to hear more of is: A. Unbiased information about how the plan is going so far -- people signed up, people calling/creating accounts, etc. B. Unbiased analysis of how many people are made better off, and by how much, and how many people are made worse off, and by how much. In other words, are we as a country better off under ACA? My guess is (1) it's too early to tell for sure right now, and (2) eventually without question, yes. But you don't hear much about A. and B. because it's hard to sell newspapers or get click-throughs with boring headlines like "An analysis of the costs and benefits of ACA."

Art my apologies for misunderstanding -- I sometimes have trouble reading descriptions like this because everyone has their own style of notating them, and I read your column headers (wrongly) as "LHO" and "RHO." If the spades are on your RIGHT, then the spade lead is unquestionably the right one in my view, given the info that you have. If partner has at least 2 spades and the J♠ or A♠ we are off to a really good start.

You cannot possibly wage a winning PR campaign against Fox News and the like, because they will find a way to twist anything into something bad for the other side. And most of its viewers are simpleton enough to take everything they said at face value. It's not a news organization, it's a political entertainment organization (with tremendously bad consequences for our current political dialogue). I agree single payer would work better than Obamacare but I believe wholeheartedly that Obamacare beats the pants off what we have now. Holding people accountable for the actual (expected value of) costs of providing healthcare to them? All the while taking money from the wealthy to pay for subsidies for the less fortunate and have EVERYONE covered? With OUR CURRENT level of income inequality? What a brilliant idea. -Tate

[hv=pc=n&s=sath75dkt972caq94&n=sq9432haqjt8dack5]133|200[/hv] Forget the bidding -- it was atrocious. I was hurricaned into the NT small slam by a know-nothing partner. My question is -- what are the odds of making the contract? How about 11 tricks? I tried for 12 and made a grand total of 10. Note: the opening lead is a small club. I realize the odds go up if a spade is led, but it wasn't.

Anything but ♠ could be right, and I lead the 10♣ because it's the minor without the Ace. Partner, if he has 5 ♥, is going to have the same problem I have in ♠ -- the LHO holds them. I should know after trick 1 whether the opening lead was right, and if I get in with a spade early I can judge where to switch if it was wrong.

This website has a rating based on hands you've played. It seems like it needs some work overall but is a pretty good indicator: http://bboskill.com/ Just type in your username and it will give you a rating. Roughly speaking, it assigns the following ratings: Below -1.4 IMPs average: NOVICE -1.4 to -0.6 IMPs average: BEGINNER -0.6 to 0 IMPs average: INTERMEDIATE 0 to 0.6 IMPs average: ADVANCED 0.6 to 1.4 IMPs average: EXPERT Above 1.4 IMPs average: WORLD CLASS I'd actually change it a bit if I were building the model, roughly as follows: Intermediate covers -0.6 to +0.6 Advanced is 0.6 to 1 or so Expert about 1-1.5 World Class 1.5+. This is because Intermediate is designed to handle the great majority of players who aren't superstars at the game but know what they are doing on most decisions in the game. Those IMP estimates in my version are just approximate based on experience. But you have to keep in mind that they are assigning these ratings strictly based upon one dimension of your performance -- your average IMP score (or MPs converted to IMPs). Given that that is the only information available, it's not a bad system, just one indicator. There are many other factors that go in to determining how you stack up. In my experience a +0.6 average is way better than average. To consistently take 60 IMPs from the opponents on average every 100 hands, day in day out, is a strong effort. It doesn't sound like a big number, but consider it this way: 100 hands takes about 8 hours to play online assuming no interruptions. That's about a work day, and if you're "earning" 60 IMPs every day you're doing very well at your job. The flip side of it is, there are a lot of very poor players on BBO, so if you're only playing against randoms all the time -- if you're leaving when the competition is decent and racking up big IMPs against beginners, flubbers and fools -- then it's not a big deal at all. If you're playing, on average, against average opponents, I'd be happy to partner you.

Hand 1: 3NT. Less than 22 HCP in my book because of the short secondary honors in ♦ and ♠. Hand 2: TOUGH. It depends on how lucky i'm feeling. Partner needs specific cards for a slam to be on in most cases. I would guess it no better than 50/50 to make slam on average. I'd probably double then go 4♥ if partner bids them or 3NT if not. And I'd grit my teeth if he showed up with something like xxxx♠ K9xxx♥ xxxx♣.

I blame the double of 4♠ first, then the failure to switch at trick 2. Partner has to have an ace somewhere and a diamond return looks threatening so I return a club. 1 down immediately. But seriously, anyone who doubles on that auction is a bad gambler.

Choice has to be between double (to show extra values) and pass. It's very likely any response by partner will lead to a minus score, and given the poor intermediates and playing MP good chance one opponent will double the final contract. Even -1 doubled is a tragedy. It all comes down to whether you think the opponents are on the level -- they may not be, but if you're wrong you go down hard. Opponent bidding suggests partner has maybe a queen and a jack somewhere.

Not a big fan of the proposed 3NT rebid, but it's gotta be at least 2NT or 3♦...

The problem with THAT is that it makes things double-dummy that are anything but. But I cannot imagine a more worthy recipient of some good fortune from bad programming than my good friend Dave! Maybe on defense, GiB takes into account that the declarer may be a first-rate class act in his dealings with people. B-)

I'd tend to open 1♠ rather than 1♣, preferring to show my 7CM rather my 3cm. ;) As far as rebids over responses of 1NT or 2 of any other suit, I jump to 3♠ in deference to my playing strength and trick taking ability in spades (or possibly NT), and my point count is at best a tertiary consideration. Since I have a play for 4♠ opposite as little as Qx♣ xxxxxx♦ xxx♥ Kx♠, and since changing the Q♣ to the A♣ makes it almost guaranteed, I think this jump to 3♠ is quite good.

Agreed. The shape is terrible and the auction unfortunate, but this is as good as a 4333 11-count can be, except for something like AK and an A and some tens. I would rebid 1nt, with an eye to shape and point count-showing, leaving the ♦ stopper or lack thereof on the sideline. If partner can follow with a ♦ stopper or club support or a 5CM, great. If not, 1NT may be our best spot, diamond stopperlessness aside.