Join Bridge Winners
All comments by Doug Bennion
You are ignoring the author of this comment. Click to temporarily show the comment.
Anybody have a link to this video #4? Thanks.
June 4
You are ignoring the author of this comment. Click to temporarily show the comment.
2 is weak, should have so specified. Dunno if 3 forcing. Not even sure if 4 is forcing although 2 bidder thought it should be.

RF: 3NT would have been down several after heart lead, diamond back.
May 26
You are ignoring the author of this comment. Click to temporarily show the comment.
Looks like the web version interface is now very similar to what the Android version has been for years.
May 24
You are ignoring the author of this comment. Click to temporarily show the comment.
My deal generator basically agrees. I asked for 50K random deals, parcelled them into 28 board sessions, and checked HCP totals for each session … 0.44% were fewer than 224 HCP (8 of 1786 sessions).
May 11
You are ignoring the author of this comment. Click to temporarily show the comment.
A couple more. At 1:28:17 FF goes to place the pad on the side-table, then rethinks it, having not actually looked yet at his hand! After seeing his hand, places the pad 1:28:45. CN appears to peer 1:28:46 and again 1:29:16.

Another. Pad placed 1:42:37, peers at 1:42:47 and 1:43:00.

Another. Pad placed 1:57:25. ‘Obvious’ peer 1:57:33, also around 1:58:00 and while chatting with his LHO.

Haven't yet seen CN peer towards the other corner of the table (his right side), all his focus is towards the side-table.
May 4
You are ignoring the author of this comment. Click to temporarily show the comment.
Jeff thanks, fixed
May 4
You are ignoring the author of this comment. Click to temporarily show the comment.
That's very interesting. First hand England match, after all the furniture arranging, and after looking at his hand, FF writes something on his pad, and places the pad on the carefully-placed table. That happens ~ 9:40. At about 10:00, CN seems to peer at the pad … it's certainly a peer in the direction of the pad.

I'm just randomly pecking here. At about 1:06:40, FF repeats … he looks at hand, he writes something, he places it on the table. At about 1:07:20, CN sneaks a furtive look at the pad.

Or maybe I'm seeing things … I'll try one more. Look for FF dropping pad at 1:13:25 and CN peering at 1:14:50.

edit: typo
Dunno.
May 3
Doug Bennion edited this comment May 4
You are ignoring the author of this comment. Click to temporarily show the comment.
On the other hand, ties are plentiful. Kibitz a BBO table with Double Dummy ON, and you'll often see lots of ties. Sometimes playing any card in the hand produces the same result. In that case, the programmer might try to make the plays look more *humanlike* (it can't cost for cosmetic fixes if all played cards are equal). For example if the bot is ‘running’ a long suit in notrump, his play might appear more humanlike if he continues playing that suit to the end, instead of jumping from suit to suit with different cashers, if that makes sense. Otherwise, the bot tosses more coins to settle ties.
May 3
You are ignoring the author of this comment. Click to temporarily show the comment.
DG is right. A tie or ties is rare. Consider a finesse with a 50% chance. The sample size of proxy deals might be 50. The finesse working *exactly* 25 times is only something like 10%. GIB would then basically toss a coin for the winner.
May 3
You are ignoring the author of this comment. Click to temporarily show the comment.
It's easy enough to program a bot to *give* a signal, it's much more difficult for them to *read* signals, and I'd be very surprised if GIB does so. Competently. I know from experience.
May 2
You are ignoring the author of this comment. Click to temporarily show the comment.
You can show both holdings if you play transfers there. X = hearts, 3 = good raise, 3 = competing.
April 29
You are ignoring the author of this comment. Click to temporarily show the comment.
If anyone overbid, it was South not North. It’s close, but the South hand relative to an average 10-pointer, is weaker than the North hand is to an average 15-pointer.

An average balanced 15 opposite 10, makes 3NT 58% of the time.

That North hand, opposite random 10 balanced, makes 3NT 61% of the time, performing better than your average balanced hand. The hand has half-an-ace more than average, the honours are working together, and 4333 is NOT the weakest balanced shape in notrump contracts … 4432 is.

That South hand, opposite random 15 balanced, makes 3NT 42% of the time. That hand is below average. It is 4432, the honours are clustered in the shorter suits, and the diamond jack isn’t working full time.
April 27
You are ignoring the author of this comment. Click to temporarily show the comment.
In Double Dummy world, It’s all on partner. If partner is in game range, that hand performs like a ‘good’ 15-pointer. However if partner has barely enough to respond to 1, then 1NT is inferior to a 2 contract.

But wait … if partner is weak, then the opps are much more likely to find their excellent major fits when the opening is 1, not so much over 1NT.

So 1NT looks like the better call …
April 13
You are ignoring the author of this comment. Click to temporarily show the comment.
PH: I used a weak notrump framework. I simmed for hands with 6+ clubs or 6+ diamonds or at least 5-5 each, any of which would probably want to act immediately. The major constraint for the comparison with invitational hands, was the HCP range limited to 11-12.
April 1
You are ignoring the author of this comment. Click to temporarily show the comment.
I'm missing something. Don't you have the same ‘directional double’ problem whenever 2 isn't natural? And my (quick and dirty) simming suggests you'll have one or both minors about twice as often as you will an invitational balanced hand, so switching up 2 and 2NT increases the problem, not decreases.
April 1
You are ignoring the author of this comment. Click to temporarily show the comment.
Thanks. That is an improvement.
March 30
You are ignoring the author of this comment. Click to temporarily show the comment.
The DONT double could vex. Opener starts 1NT, overcaller doubles playing DONT, which ostensibly shows a long suit (advancer usually bids 2, P/C), but is convertible by advancer. Responder has a bad hand with which he would normally run, but hopes the double has let him off the hook, as it usually will. Now advancer passes. What should opener do, knowing that responder’s hand can vary from complete bust, to a nice 9 or 10 or so?

Haven’t had it happen to us yet, but I think if it ever does, I’m running as opener if I have a 5-card suit.
March 30
You are ignoring the author of this comment. Click to temporarily show the comment.
Transfers beginning with XX. A transfer cuebid is whatever your cuebid would show. xNT is natural, having positional value. For us, (1) 1 (X), then

XX = clubs, here the transfer-cue, showing a limit raise or better
1NT = natural
2 = diamonds
2 = hearts
2 = constructive 3-fit raise
2 = courtesy raise

3 = mixed raise
3 = preempt
March 30
You are ignoring the author of this comment. Click to temporarily show the comment.
We use 2 = size ask or m, and 2NT = mm. Works fine but any time you have a dual-purpose bid, you're potentially subject to preemption.
March 28
You are ignoring the author of this comment. Click to temporarily show the comment.
If the opening is 1X instead of 1NT, overcaller has far more actionable hand-types. Say RHO chooses to open 1 rather than 1NT. 1 opens up (1) lightish 2m overcalls (most 1NT defences use 2m overcalls for nefarious unnatural reasons), and (2) takeout doubles of 1, the likes of 1-4-4-4 and 1-(543) and 2-(443) etc, hands often with no suitable action over a 1NT opener.

I ‘looked’ at that position once (when wondering if I should open 5M332 in a weak notrump), and overcaller owns something like 30-40% more actionable hands when RHO opens 1 instead of 1NT.
March 19
.

Bottom Home Top