USBC small slams

[vsp...@hotmail.com]

USBC small slams

5 times a small slam was bid a one table and
a grand with bid on the other.
4 times the small slam made and the grand failed.
One time both made.
Score
small slam 64 imps
grand slam 11 imps
---
23 other times a small slam was bid on one or both tables
3 times bid and made on both tables
one time bid and failed on both tables
one time different slams bid; one made and one failed
18 times slam bid on only one table
9 times slam made. +108 imps
9 times slam failed. -100 imps

From these 23 plays
17 of 28 times small slam was bid and made.
Plus the boards where grand was bid
the totals are 22 or 33 small slam bid and made.
3 of 10 grand bid and made.
Doesn't seem like a very good record for the
best players in the U.S.A. The main reason the slams
failed were lack of HCP, no source of tricks. Mostly
it was just plain overbidding.
On over 10% of the boards one of the four pairs
contracted slam. It only made on about 6% of those
boards. These online forums seems to contain an
inordinate number of threads on slam bidding. Bridge
players seem to have an abnormal fear of failing to bid
a makeable slam. Hand histories show that many more
imps are lost bidding fathom unmakeable slams.

jogs

[Stu Goodgold]

Interesting stats. Can you tell us of the 5 grands bid with 4 failing, how many were made by the side that was trailing and needed a possible swing?
It may be a case of desperation rather than overbidding.

-Stu Goodgold
San Jose, CA

[vsp...@hotmail.com]

Diamond vs Lee
60. & 68. Both grands bid by Lee.

Nickell vs Milner
60. Grand by Milner.

Nickell vs Diamond
54. Grand by Diamond

All four times the grand was bid by the
losing team. Diamond was ahead of Nickell
at the time that grand was bid.

[rhm]

I am not sure your conclusions hold.
The sample on grands is quite small and may not be significant.
You only need slightly above average odds at IMPs for the grand to be
worthwhile, when the other room will be in a making small slam.
A sample of 5 is too small.

With regard to small slams you require about a 50% chance (single
dummy) of making it at IMPs.
Some but not many deals will of course be much better or even lay-down
for 12 tricks with no good chance for 13 tricks. .
So you have to expect that quite a few small slams will fail at the
table or you are bidding too few worthwhile slams.
When 17 out of 28 slams actually made, your conclusion that "Doesn't

seem like a very good record for the

best players in the U.S.A." does not necessarily follow from these
statistics.
A success rate of 60% is certainly nothing to write home about, but a
success rate of around 75% is probably the best you can accomplish.
So 21 out 28 would probably mean near perfect bidding.
Of course inspecting actual deals may tell a different story and
support your conclusions.

Rainer Herrmann

[vsp...@hotmail.com]

While 240 boards with 330 matchups is indeed
a small sample, the slam bidding frequency is
consistent with what has been published by
other authors.
About 10% of the time one of the four pairs
on the table will bid slam. Less than 5% of
the time does one of the four pairs hold 30+
HCP. Of the 8 small slams that failed only
one was because of a losing finesse. The other
7 had no play. Missing two aces. Missing ace
and king in a suit. Poor trumps. Fewer than
25 HCP. Mostly the pair had no business being
in slam.
With grand slams if your team is bidding more
than one grand every 100 boards, neither one
are very lucky or more likely you guys are
overbidding.

[vsp...@hotmail.com]

On Wednesday, May 16, 2012 3:35:14 AM UTC-7, rhm wrote:

The 8 slams which failed were all in the
semi finals.
boards 8,41,54,58,81,88,112, and 119

[vsp...@hotmail.com]

Let's treat slam bidding as a statistical hypothesis test.
The first step is to state the relevant null and alternative hypotheses.
The null hypothesis is a slam exist on this board.
The alternative hypothesis is no is no slam on this board.
Type I error: not bidding a biddible and makeable slam.
Type II error: bidding an unmakeable or low percentage slam.

The evidence from results of play by the experts is that they playing to avoid type I error. Therefore they make type II error far too often. Most experts would be better off just missing every slam whenever their partnership HCP total is 26 or fewer. They would be gaining imps using this passive style.