Join Bridge Winners
All comments by Stefan Olausson
1 2 3 4 5 6 7 8 9 ... 43 44 45 46
You are ignoring the author of this comment. Click to temporarily show the comment.
If you rebid 1nt with 4sp balanced, you usu lose when responder is weak with 4spades.

A compensating factor is that you rightside the 1NT contract, when responder is weak *without* 4spades (more frequent).
The lead & defence also becomes trickier when they dont know declarer's spade holding.

Missing the 44 spade suit, probably is more costly at MP than IMPs.

Anyone's choice which method/style they prefer.
June 22
You are ignoring the author of this comment. Click to temporarily show the comment.
Yes, at St Erik we often have international guests playing.
June 22
You are ignoring the author of this comment. Click to temporarily show the comment.
From 30 July (not June), perhaps you mean?
It starts july 26.

The std system in sweden today is pretty much 5542 2/1gf, like “bbo std”, except 1M-1NT is usually non-forc. .
June 22
You are ignoring the author of this comment. Click to temporarily show the comment.
Deleted. Misread.
June 21
You are ignoring the author of this comment. Click to temporarily show the comment.
“so it looks like we'll make the contract either when trumps come home (50%) or when East has both club honors (26%)–63% in all.”

Also add a few % for east having Hx.
June 16
You are ignoring the author of this comment. Click to temporarily show the comment.
You forget to mention which contracts we are in.
June 16
Stefan Olausson edited this comment June 16
You are ignoring the author of this comment. Click to temporarily show the comment.
Even tougher, what do you do with these
a) xxx Jxx Kxxx AQx
b) xxx Jxx Axxx AQx
after 1-(1)?
June 16
Stefan Olausson edited this comment June 16
You are ignoring the author of this comment. Click to temporarily show the comment.
“Which hands that contain fewer than 4-card ♥ would double 1♠?”

Maybe one of those?
a) xxx KQx Kxxx Jxx
b) xxx KQx Kxxx Qxx
c) xxx KQx Axxx Qxx

Not ideal for NT without sp-stop, a bit too strong for pass(?), neither 2 (vs 2+ opener), nor 2 seem attractive.

Personally, I prefer X guarantees 4s, because these “awkward” hands are rare in practice, but not sure what I would do with the above hand – probably X anyways as the “least evil”.
June 16
Stefan Olausson edited this comment June 16
You are ignoring the author of this comment. Click to temporarily show the comment.
This was the actual deal:

https://tinyurl.com/3H-lead
June 15
You are ignoring the author of this comment. Click to temporarily show the comment.
Re. Option 1 — why would I ever double with this hand?
What is that supposed to mean?
June 14
You are ignoring the author of this comment. Click to temporarily show the comment.
I particularly detest the jump to 3 with a semi-balanced hand and tolerance for either 5+major.
It's a bid that should generally be avoided, because it wastes so much useful space and gets in the way of the strong opener.

Start 2-2-2-3, and should not be hard to find the heart-slam.
June 9
Stefan Olausson edited this comment June 9
You are ignoring the author of this comment. Click to temporarily show the comment.
hehe… true, it's a tricky area :)
June 6
You are ignoring the author of this comment. Click to temporarily show the comment.
re “ethics side of bidding” my view is that it does not exists in robot-tourneys.
If you think you found a way to explore robot-weaknesses, it's just part of the game.

I really don't know if the above example (concealing the long suit) is a winner or loser in the long run. I would guess loser, but maybe not.
June 6
Stefan Olausson edited this comment June 6
You are ignoring the author of this comment. Click to temporarily show the comment.
At the time it was developed (mid 1990'ies), GIB algorithm was a great breakthrough over all earlier attempts at programming bridge-player software.
But after the “Gibson” single-dummy analyser (early 2000's?) virtually nothing has been done to improve.
June 6
You are ignoring the author of this comment. Click to temporarily show the comment.
Yes, GIB robots are very weak in many defensive situations
because they have virtually no defensive signals or agreements.

Not only that, they also draw no conclusions from earlier play.

I had a spectacular example yesterday, I was declarer, and during the play RHO-bot played a small heart from AQx, I had KJ alone on hand, and won the Jack.
Later, at trick 12 RHO needed to discard from: - A - A.

A human would of course conclude that since the J won, declarer must have K left, but RHO-bot discarded A “thinking” it's equally possible that declarer has K left.

Re “good programming?”
it's just that the algorithms GIB was built on are completely void of “thinking”.

It randomises the remaining cards and selects a card based on double-dummy analysis of the random-samples, that's basically it.
June 6
Stefan Olausson edited this comment June 6
You are ignoring the author of this comment. Click to temporarily show the comment.
From David Gold's video-link above, if the play of the club-jack is the one referred to, certainly a hair-raising conclusion that “the play was made in tempo”.
June 4
You are ignoring the author of this comment. Click to temporarily show the comment.
>“…helping determine who is at fault for slow play…”

This is also an interesting point, I think.

Just wondering, if a match (or half of a match) of, say, some 16+boards, exceeds the time-limit, it seems like a very tedious task to go through a whole video and sum up how much time each pair used at the table.

Does that really work in practice?
June 4
You are ignoring the author of this comment. Click to temporarily show the comment.
@Martin Lindfors

And still, how can responder know what to “expect” if the Double can contain both a stack of spade tricks, as well as a void in spade?

Would you Double with both hands above?
And what if you subtract a King from both hands?
June 2
You are ignoring the author of this comment. Click to temporarily show the comment.
“Double is penalty, otherwise opponents can preempt 4♠ with any junk and not risk much.”

In practice, no.
If you play t/o, both players can still X and pard converts.
June 2
You are ignoring the author of this comment. Click to temporarily show the comment.
“Cards”?

So you can have anything from, say
AQTx Axx Kxx Axx
to
- AQxxx KQxx AKxx
?

That seems like a horrible agreement, unless you combine with
quick-X=penalties, slow-X=T/O, or similar.
June 2
Stefan Olausson edited this comment June 2
1 2 3 4 5 6 7 8 9 ... 43 44 45 46
.

Bottom Home Top