Tuesday, January 26, 2010

Swiss teams puzzle

If it takes 30 seconds to shuffle, and 30 seconds to deal, how long does it take 2 people to make 3 boards at the start of a swiss match?

Answer below...

For most people, 2 minutes. Both people shuffle and deal a board, which takes 1 minute, then one of them makes the other board which takes another minute.

But, by scheduling better it only takes 90 seconds. Call the players A & B. Then:

0-30: A shuffles 1, B shuffles 2
30-60: A deals 1, B shuffles 3
60-90: A deals 2, B deals 3.

My partner is very slow, so I have to think about these things.

Read more!

Tuesday, January 12, 2010

Skip bid

In order to better convey the impression that I'm actually thinking after a skip bid, and also to introduce random variation in the length of a hesitation, I (when I don't actually have anything to think about) add up the spots in my hand.

That's all.

Read more!

Tuesday, January 5, 2010

After 2nt rebid

Over at Becker's Bridge Blog he posts his system after 1m-1M-2N.

I play a sort of hybrid between transfers and what he describes:

(Also applies only over 1m-1M-2N)

3C is a relay to 3D. Might be left there. Any subsequent bid, though, show opener's minor:

3N = mild slam try with 4m
3M = 5M, 4m
3oM = 4M, 5m

3D is a xfer to hearts (opener should break with preference for spades after 1m-1S-2N-3D)

3H is a xfer to spades (after 1m-1H-2N-3H, responder could be 4=4 or 4=5 in the majors, so opener bids 3N with support for neither, 3S with 3 *hearts* and past 3N with 4 spades)

3S shows the other minor

4C RKC for responder's suit

4D shows 6 of the other minor

I think showing some good supporting hands below 3N is useful, but I also like his idea of 1m-1M-2N-3C-3D-3M as showing a 6+M slam try. My style requires going beyond 3N with that hand, though that doesn't strike me as a big cost. On the other hand, I think it's more important to suggest opener's minor and still be able to stop in 3N; Becker's system doesn't seem to have that sequence at all.

Read more!

From the Reisinger

I watched this hand on Vugraph, deal 20 from the first final session of the 2009 Reisinger.

10 9 5
8 6 2
A J 10 5
Q 5 3
K 8 2
A Q J 7 3
A K 9 6

West North East South
PassPass2♠ 3
All Pass

Lead: ♠Q

The panel had a strong consensus that North is not supposed to raise here. I agree.

After some hesitation, East ducked the lead to declarer's K. Declarer crossed to A and lost a heart finesse. A diamond came back, ducked in dummy, East played the K and declarer ruffed. After drawing a 2nd round of trumps (both following, East with the 10), declarer has 9 tricks and is left to try for a 10th.

Declarer left the last trump outstanding and successfully ruffed a club, finding West with 1=3=5=4 distribution.

This seems pretty likely to work, but I wonder if it's possible to do better. As I write this, I haven't done the analysis, so don't be prejudiced by my presentation. An analysis of an alternate line is below the fold...

This is the end position:

10 9
J 10
Q 5 3
8 2
J 7
A K 9 6

I think we can treat it as certain that spades started 1=6. Playing clubs as declarer did succeeds when clubs are 3-3, or when long clubs are with the remaining heart -- this is only possible with West being 1=3=5=4 or 1=3=4=5 (or, very remotely and henceforth ignored, 1=3=3=6). I'm also ignoring whatever inference is available from East's failure to insert the 9.

The diamond position creates some interesting loser-on-loser possibilities, even though short a dummy entry.

My first alternate line was to draw the last trump and exit a spade, aiming for a minor suit squeeze. If the defense wins 2 spades, you're set up and just need the squeeze to operate. If they win 1 and play a diamond, you can just pitch your losing spade and you'll get a diamond trick now or later. But, East can thwart this by winning the spade and playing a club. You can still try ♣A, ♣Q, and try to throw West in with a diamond. This will work if clubs are 3-3, or if East has short clubs but at least 1 honor and West has the Q.

This is too much work, so I'm going to start by assuming clubs are 4=2 (everyone makes when they're 3-3). Restricted choice applies to both the Q and the 9. If East is 6=3=2=2 there are 3 possible holdings (6 small diamonds, 1/2 for restricted choice in hearts). 6=2=3=2 without Q is 15 (6c2 diamonds). 6=2=3=2 with Q is 3 again.

Declarer goes down in the first case, but makes in the other 2 (18 total). Meanwhile, with 2 clubs East has at least one honor 60% of the time, so my alternate line will make the overtrick in 60% of the first 2 cases (but fail in the 3rd). That's only 10.8 total. Misdefense is plausible but not plausible enough to make up this difference.

A different alternate line is to draw the last trump, play 3 clubs ending in dummy, then pitch a spade on a diamond honor. If clubs are 3-3, this makes 11 tricks (West will be diamond tight and must have Q). If clubs are 4=2, this makes 10 tricks when West has Q, 9 otherwise (same odds as declarer's line, though different losing cases). When clubs are 5=1 this only makes 9 tricks (unless East has a stiff honor, then change tacks and win the 2nd club in dummy). Of course, declarer at the table also is held to 9 tricks when East has the 3rd heart with a stiff club.

This is much closer than my first idea. Comes out ahead when clubs are 3-3, 20 positions x 6 diamond positions = 120. Comes out behind when East is 6=2=4=1 with a small stiff club (4) x (20 Kxxx diamonds + 7.5 KQxx diamonds) or a stiff club honor (2) x 7.5 KQxx diamonds (never mind that AJxxxx/Tx/KQxx/J might open 1S). That's 125 positions.

So, declarer still comes out slightly ahead (125:120 in the positions that matter), but it's very close. If the offense hadn't held the 8 it would have swung the other way (since East 6=3=2=2 would be twice as likely without restricted choice applying to the 10.

Read more!